Spectral Unveils the SDR-4000SL Master CD Processor

Spectral Audio , the northern California company whose director of engineering is Prof. Keith O. Johnson, gave the first public preview of its SDR-4000SL Master CD Processor on September 24, at Music Lovers Audio, in Berkeley. Introduced by Johnson and Spectral founder Richard (Rick) Fryer, the $19,000 Spectral Digital Resolution (SDR) model sounded sensational playing 16-bit/44.1kHz, HDCD-encoded files Johnson had made for Reference Recordings, through a system that included Spectral's DMC-30SS preamplifier and monoblock amplifiers, Wilson Audio MAXX 3 speakers, Spectral Ultralink II speaker cables, the MIT Z Duplex conditioner, Synergistic Research's controversial ART system, and other room treatments.

Fryer, who cut short his graduate studies in psychology to found Spectral Audio in 1977, when he was 23, noted that his company's marketing style is decidedly low-key. No wonder: demand for Spectral products is so high, and production so limited, that only 900 units of the SDR-4000SL's predecessor, were manufactured between 2003 and 2011. Samples of the SDR-4000SL have yet to reach most of the company's dealers in the US and abroad, but I was told that already the waiting list for it exceeds the initial production run.

"Don't count 44.1kHz and CDs out," Fryer proclaimed at the start of a lengthy, technically complex introduction that he delivered without a single hesitation. "If 44.1 is implemented well, it is high-resolution. The problem is, there is little in the high-end digital arena that really is very good. CDs have an inherent nature all their own that needs to be supported by the finest technology they were designed for. If you want to hear the best that 16/44.1 can deliver, you have to revisit the technology, as we have, and develop a one-piece, single-box solution dedicated solely to the CD."

Fryer and Johnson set about to merge their finest transport and playback technologies into a single, "price has never been an obstacle" box. Separating CD-player technology into its six or seven subsystems, Fryer said, they "turned over every stone" to build, from the ground up, a one-piece component that would deliver sound similar to that of the legendary Pacific Microsonics Model 2 processor.

Among the weaknesses of traditional CD players that they addressed were clock-induced jitter and the limitations of op-amps and integrated circuits. Favoring discrete circuits that foster "continuity and brilliance," Johnson spent years designing a complex DAC interface-actually, an aggregate of eight separate amplifiers. This architecture is designed to present the DAC with an unmatched conversion environment, and to provide exact waveforms that are then processed by passive analog filters. Johnson stated that, unlike integrated circuits, his discrete circuit's internal state never changes-a feature that, Fryer claims, sounds like analog. Other dedicated discrete circuits are applied to Spectral's reference DMC-30SS preamplifier and the SDR-4000SL's line section.

The Spectral team then evaluated dozens of transports, before partnering with an Asian company to modify one of their models. The result is what Spectral claims is an "exceptionally silent" or stealth transport that is slaved directly from the DAC. Signal paths are fully balanced and isolated, resulting in clock-derived buffered signals with apparently no trace of rotation, vibration, servo, or other mechanically or electrically sourced artifacts that can cause jitter and thus degrade signal conversion. Gone are belts, isolators, or rubber parts, which degrade over time.

"Our machine reproduces many traditional test signals more accurately than commercial test equipment," Johnson claimed. "Because the smallest amount of noise will mess things up, we minimize accessories, and slave everything in the DAC to a dedicated fixed-frequency crystal. We also use floating, battery-like power to isolate most parts of the machine from one another. Given this silent conversion environment, where everything functions in analog class-A, we end up making manufacturers' parts work and sound better than they ever imagined possible."

Johnson and Fryer say that Spectral will "someday" develop a high-resolution file player. Meanwhile, they're convinced that the SDR-4000SL Master CD Processor functions at such a high level that there is little difference between the sound of well-recorded CDs through their player and the best current playback devices for higher-resolution files.

After almost an hour of explication, Johnson played excerpts from a few 16/44.1, HDCD-encoded tracks from the Reference Recordings catalog. I was so blown away by the extremely long decay time of the triangle in Benjamin Britten's Young Person's Guide to the Orchestra (Michael Stern, Kansas City Symphony, RR-120 CD) that I asked Johnson for an explanation. He replied that while "Red Book" CDs lack enough bits to reproduce such die-offs in their entireties, HDCD's code adds a bit or two that sustains them to their natural ends.

If the triangle pricked up my ears, the unexpectedly huge drum thwacks in an excerpt from Britten's Sinfonia de Requiem, from the same disc, were so startling that I literally jumped a bit in my seat. (I was seated in the first row, somewhat in line with the left-channel Wilson MAXX 3.) As Johnson played a few more examples, it was easy to hear the difference between the drier sound of miking in a large arena with virtually no early reflections, and the wetter sound of swing music captured in a small hall. By the end of the Reference Recordings portion of the demo, I was certain that, in terms of dynamic range, dynamics, smoothness, and control from top to bottom, Johnson, Fryer, and the folks at Spectral and Music Lovers Audio had treated us to the most sensational demo I'd ever heard. After the Spectral men had left the demo room to schmooze elsewhere, Music Lovers' Hugh Fountain played some more selections. Burns of older CD tracks to CD-Rs, as well as a superior remastering of a classic Santana album, further confirmed that the Spectral SDR-4000SL Master CD Processor was taking no prisoners and telling no lies. If the CD it played had problems, we heard them. But when the recording and pressing were superb, we were rewarded with far more of their musical bounty than is the norm.

comp.audiophile's picture

Hi Jason - Is this the same CD player that Maier used back in June at The Show Newport? I know the one he used was "New."

Skeptica's picture

$19,000 for a CD Player? And no ABX test to see if there is any improvement over a standard CD player. Let's keep in mind that the distortion of speakers and the room is typically many orders higher than that of a typical CD player.

It is true that there is "little difference between the sound of well-recorded CDs through their player and the best current playback devices for higher-resolution files." But that is true for every decent CD player. The reason high-resolution files typically sound better is because very well-recorded music is usually selected for the high resolution encoding treatment. Scientific studies have shown that resolution higher than CD's 16-bit/ 44.1 KHz is not audible to pools of listeners. Most music is not recorded all that well for various reasons. A CD player isn't going to make a difference. But if you want something unique that your neighbor doesn't have then this might be a player to get.

John Atkinson's picture

Scientific studies have shown that resolution higher than CD's 16-bit/ 44.1 KHz is not audible to pools of listeners.

There has only been one such study of which I am aware, by Brad Meyer and David Moran in the Journal of the AES. This study was methodologically flawed.

John Atkinson

Editor, Stereophile

ralphf's picture

While it may be true that there is "little difference between the sound of well-recorded CDs through their player and the best current playback devices for higher-resolution files."  I would imagine that is easier and therefore less expensive to achieve that result with more data available. 

volvic's picture

I can only speak from experience but I have always coveted their CD players.  Long long time ago say around 1991 I heard a SDR-1000 with a CD I was very familiar with and was left speechless - at the time I thought it could not get any better than a Karik/Numerik combo and boy was I wrong.  Shortly after that I acquired a Karik/Numerik combo and was still left wanting.  How could it sound so much better?  Yet it does.  I can only imagine how good this one is, sadly very few of us will ever hear it or own one. 


Stephen Scharf's picture

Despite what Fryer says about not counting CD's out, for better or worse, CD players are out. Music servers are in. Why continue to spend significant R&D on developing expensive transport mechanisms and costs that are passed on to the end user when you can eliminate the transport altogether from the equation? The only purpose I see for CDs now (and there are some excellent ones, notably the Audio Wave XRCD24s from Elusive) is to be ripped for playback via a music server. Couple a good content storage drive, player software, and an excellent asynchronous DAC to a computer, and one can obtain extremely high levels of playback quality now for the a fraction of the price of the Spectral, as well as having the ability to play the ever-increasing library of high-res files. Make no bones about it, 24/96, 24/176 and 24/192 sound markedly better than the same content on Redbook. Still not as good as LPs, but close enough to create an engaging experience. 

Saying CD players are still a reference is like saying Tri-X is a reference for image quality when there are digital cameras like the Canon 1DX available. 

Dr. John Gotwalt's picture

Is the performance of this player optimized in an all-Spectral system, or does it play well with others ?

A_Stone's picture

I have to, reluctantly, agree with "Skeptica" and his comments on the subject of perceived sound quality of CD vs high resolution 24 bit xx files.

After conducting very extensive research on the subject, it is now clear to me that, for the mastering process, the recording "quality" is extremely important for the final result. A poorly conducted recording will not yield a better result once quantized to 24 bits 192 kHz as opposed to 16 bits 44 kHz.

For the playback process, the higher resolution signal, at least theoretically, is capable of recreating the original signal closer to reality IF all other parameters involved in the recreational process are kept unchanged. (Performing blind A/B tests showed that our test panel - audiophiles and non audiophiles - were all UNABLE to conclusively find definitive differences between the same audio source playbacked at 16-44k or 24-192k).

Here is were High-Resolution signals run into problems. Jitter becomes much more difficult to tame for the higher bit rates than that of CD clock frequencies, and the human auditory system seems to be more sensetive to timing imprecision inside playback devices (time-axis) than to the amplitude imprecision of the regenerated signal (amplitude-axis).

If you havent tried this, just find a jitter-filter (audio-alchemy dti or genesis d-lens), connect the device between your transport and DAC of choice, and see (hear) the shocking result for yourself.

As long as the logic building-blocks are flank-triggered, jitter remains a compromise. Any noise coupled to and from the peripheral devices only accentuates the jitter problem: At higher bit-rates these are (noise & jitter) far more difficult engineering challenges for signal fidelity than simply increasing the quantisation and sample rate of the quantized signal.

with regards: A_Stone