by David Lander
Though Spectral Audio has been producing audio components for more than two decades, readers of U.S. hi -fl magazines may find their mental picture of the company a bit ghost-like. Limited press coverage in this country is almost certainly the reason. American writers, Spectral founder and owner Richard Fryer explains, have been reluctant to review entire systems. Yet that's precisely the way Fryer says his products need to be tested. They are, he emphasizes, designed to perform together and can be judged fairly only as an ensemble.
The son of a physicist doing government work, Fryer was born in Los Alamos, New Mexico, in 1953. Soon after, the family moved to California, where Professor Edward Fryer's work--Rick's father taught at Stanford University and Pomona College--led to his son's spending a good deal of time on and around college campuses. The milieu, one in which a lot of questions are asked and ideas shared, had a lasting effect on him, Rick Fryer reasons.
-----Richard Fryer (right) and Keith Johnson, Spectral's Director
of Engineering
Fryer's route to the high-end audio summit led him to jobs in high-end retailing (at a Los Angeles shop called 2001 Sound Odyssey) and manufacturing (with speaker maker ESS). He also worked in publishing, at a small audiophile magazine, Sound Advice, where he established test procedures and reviewed equipment. (And where he worked with Damien Martin, who was to become a partner in Spectral and its first engineer.) Eventually, Fryer found his way to Northern California's high technology epicenter, Silicon Valley, where he put down corporate roots and continued his quest for the ultimate in music-reproducing techniques.
As for the company name, it has nothing whatsoever to do with ectoplasm or phantasmagoria. Spectral, Fryer explains, is a reference to the spectrum from DC to light, an allusion to the extremely wide bandwidth of the products that wear the brand. It's also meant to suggest the lucency and uncountable colors of music, which Rick Fryer has devoted his career to reproducing for serious listeners.
-D.L.
I've heard that when you were working on prototypes for what would become the MS-1 preamp, the first component to carry the Spectral name, you made a walkabout into other fields to find out about more advanced technologies than those being used in the audio industry. Tell me a bit about that.
It exposed me to the cutting edge. There's a considerable difference between what is done in fields like radio, communications, and microwave-and computers certainly--and what is done even in what we would consider perfectionist audio. High-technology people, engineers, will tell you that there can be a 10-year trickle-down effect before processes, components, and technologies developed for state-of-the art areas begin to make their appearance in consumer electronics, mostly for two reasons. Truly, cost constraints are operating here; in many of these technologies, getting the performance is the issue, not the price. The other reason is simply familiarity.
And you didn't want to wait 10 years.
I felt that some of these techniques--certainly not all, but some--begged to be used in high-end audio design.
How did you choose to use them?
We started with a preamplifier design that we could take to an instrumentation level, and we developed that design with the best possible discrete parts. We did this in a facility that was tremendously rich in instrumentation capability, in an extremely successful Bay Area computer firm.
What was it called?
Four Phase Systems. The company's gone now, but they were building advanced process-control computers--industrial computers. In the '70s, companies such as Hewlett-Packard and Tektronix were developing some of their finest instruments. We were very interested in applying some of these instruments and measurement techniques. When we re viewed products at Sound Advice, we were fascinated that the great preamplifiers of the day sounded nothing alike. Bill Johnson had a marvelous achievement in his SP-2 and SP-3 preamps, and Mark Levinson and his engineer, John Curl, had a breakthrough product in the JC-2 preamplifier. But there was no commonality in terms of the sonic and aesthetic experience of these high-end products. That raises several questions. What is accuracy? What's the truth here? Is it something in between, or is it something else entirely?
I think, at first, our motivation was purely scientific curiosity. Because without having something better, something as a reference, we couldn't get to the truth. We needed something superior, so we developed a reference preamplifier that was not subject to commercial constraints, that we felt served as a better reference than commercially available products.
At first we didn't see it as commercial; we treated it as test instrumentation. If a resistor cost $4, that's what it cost. We knew the precision was higher; we knew it was better. We could get better rise time, and it was quieter.
How often did your experimentation get you closer to what you considered sonic truth?
Well, many times it didn't. We went through tremendous periods of experimentation. We threw every thing against the wall to see if it would stick, so to speak. At first, we weren't sure what the sonic consequences would be. No one had attempted this, so there wasn't a body of information. But we were pretty confident that our work would yield significant results.
Since the MS-1 cost about $1,600 at a time when other high-end preamps were in the $650 to $850 range, the first of those results proved expensive.
It did. Embarrassingly so. This was not our style and not what we wanted, but, at the end of the day, you added up the parts and the processes, and it was an expensive, hand -built instrument.
But people bought it.
Yes. More than we imagined. The dealers found out about the MS -1 very quickly, and it launched the company. We have to credit a couple of other companies for having developed dealers who could hear the difference and did care. All of this is a tree falling in the forest if people don't care.
What specific aspects of reproduced music do you think listeners should care about? What sonic at tributes concern you most?
Things like transparency. Spatial things. Resolving the detail, large and small, of a recording. Listening involvement and excitement and communication in a recorded performance tend to recede or become more subtle when they're not absolutely optimized.
There's something that says, "I'm not really here. I'm not sound occupying a real space." But when you listen to a microphone feed, many times you get this electrifying feeling of being there. We asked our selves, why isn't that sense of being there captured more completely in high-end systems?
What did your research lead you to conclude?
Some very basic things have come out. As well as linearity and other conventional factors that audio designers talk about, the factors that make these subtle and involving things occur are essentially related to speed and settling time. To be able to reproduce a loud bang and, suddenly, after that, have no memory of events, just a settling to absolute silence as if nothing had happened, you need circuits that are extremely fast. Until you hear the phenomenon demonstrated, you don't always make the tie-in.
But from the turn of the century, and particularly through the War years, there was quite a bit of re search done at the major American universities the military funded much of the experimentation that clearly confirmed many phase -related relationships. The initiation response in the ear's hair cells is at least 10 times faster than the bandwidth being sent through the middle ear, meaning that the ear is sensitive to some form of wavefront or initialization information that can be called phase or time -arrival. We need very fast circuits that settle very quickly and completely from a musical event; you don't want a signal artifact that remains after the event. That's why we need the speed.
You've said that Spectral does more volume in Europe than in the United States.
Yes. You can argue that there is a sensitivity and refinement in music appreciation in Europe, where the bigger -hammer approach is not as respected as it is in some other markets. The idea that bigger is better is not always espoused in more sophisticated urban areas. That does dovetail with what we find interesting. On some levels, we develop products to satisfy our own sensitivities. Having been around high -end equipment a good deal of our lives, we're ready for something that shows refinement and musical capability. And this doesn't come from a bigger hammer; it comes from the ability to satisfy the ear, and the ear is voracious for detail and the sense of place. We want to provide as much information, in a linear fashion, as is possible to retrieve from a recording. Nothing less and nothing more. This requires very high-performance circuitry capable of great subtlety. It does not mean big; it means, when you need the resolving power, it's there.
----- Spectral's DMC-20 Reference preamplifier and its modular circuit
board
-----The SDR-2000 Reference D/A converter
You also insist on using discrete circuitry much of the time. Why?
For critical signal applications, we've found over the years that integrated circuits simply don't meet our quality standards, although we're constantly evaluating new parts and, to everyone's credit, they're get ting better and better. Still, in the rigorous evaluations that we do, they simply can't pass the microphone feed accurately. There's so much musical information and life that's lost, even with the most premium integrated-circuit amplifiers, that ICs are not up to our needs in a critical signal application. Nevertheless, there are applications for servos and for less critical parts of the signal path where these can be perfectly workable.
Another of Spectral's defining traits is that you emphasize a systems approach that extends as far as the cable. How did this develop?
Spectral, during the early '80s, had already come up with the concepts of fast -settling, high-speed amplifiers and preamplifiers, but the interfaces, so to speak, were at a level of crudeness that did not allow us to push ahead. We met Bruce Brisson, then a consultant for Monster Cable, and engaged him to help us with this technical problem [an arrangement that led to the founding of Musical Interface Technologies, commonly called MIT]. Basically, like microwave designers or people involved in radio and high-speed analog signal work, we needed tuned systems. A cable that was more than a wire, one that was actually terminated as a low-pass filter, was a key breakthrough that allowed high-speed amplifier circuits to be stable and practical for home audio use. By defining the cable as a low-pass filter, Bruce was able to achieve a level of transient resolution-transient accuracy that was simply unknown before. With this cable technology, we can produce very-high-frequency square waves with remarkable fidelity and achieve things that amplifiers couldn't do in the past in terms of rise and settling times. It opens the door to a degree of transparency, at both low and high levels, that isn't possible with slower circuits. It does, however, require a commitment to full system integration, something that historically hasn't been done much in audio, even high-end. In the technology fields, there's never any argument about how instruments can be used together and optimized. You rarely think of a measurement system or a high-performance microwave sys tem without looking at the component parts and how they integrate and work together. In component audio, we're clearly endorsing that, too, but typically the industry falls short in the optimization. Why? Well, a lot of it probably comes down to commercial need; certainly many audiophiles derive tremendous satisfaction from mixing and matching components themselves. The only problem with this is, as we get into high-performance audio products, these designs are intrinsically complex and the interactions subtle and sonically significant. Mismatching of high-end components is, unfortunately, all too common.
You maintain a close relationship with Reference Recordings. In your view, it appears, the systems approach goes beyond products and extends to companies that work together.
It does, because no single audio company alone can make an assault this wide on the state of the art with out having collaborators. We couldn't have this successful systems approach without MIT, which developed the cable solutions that allowed us to pursue very-high-speed applications. And we could not evaluate these circuits for their musical accuracy without fundamental experience in making recordings and comparing the sound to the microphone feed. Those things are critical to a product that addresses the issue of what accuracy in music reproduction is.
Have you never been tempted to put Spectral into the speaker business?
People who have followed us have regularly seen Spectral speaker prototypes in our displays, but these are to prompt and to excite. It's very important that we understand there's no one ideal speaker, there's no one ideal speaker technology. If we could have a wish for these transducers, it would be that they be fast, articulate, well -settling, well-behaved, and powerful, so that subtlety, transparency, and spatiality are resolved from a good recording. Whether it's a big speaker or a little speaker, a planar or a ribbon, there's the capability to do this. We would not be able to address all those technologies, and we're very interested in supporting a diversity.
I gather Damien Martin, your first engineer, is no longer involved in Spectral.
Not for many years. He's involved in other audio products. He came out of school as an electrical engineer and also was a very talented film student. His contributions to the field give you some sense of how engineers can be now, that they can have a humanistic background and not be merely meter readers. The New Age audio engineer is someone who has to bridge a number of philosophies and areas of specialization.
--- SDR-3000 Reference CD transport
That seems to describe Keith Johnson, who now does your engineering. When did his involvement with Spectral begin?
Keith has been involved for many, many years. When Damien Martin started his own company in the mid-'80s, Keith assumed full engineering control. I'd known him for years. He lived in Southern California and was one of these dear people that many of my friends knew. He became involved with Tam Henderson at Reference Recordings after RR had produced maybe two recordings. That relationship with Reference Recordings exists to this day and is a foundation for our development technology.
--------Fryer's emphasis on discrete circuitry is based on exhaustive component
testing.
Please say a bit more about that aspect of developing your products. You have mentioned the importance of the microphone feed.
When there's so little consensus as to accuracy in these state-of-the-art products--every system you hear is quite different-and when you interpose one of these components or systems between you and a microphone feed and it changes the sound substantially, you have to ask, what is this doing to a recording? There's tremendous tuning that goes on, not only in audio components--preamps, power amps, turntables, etc.--but in cables and accessories. All of these can change the sound, and change isn't intrinsically good. You have to ask yourself, when is it more completely faithful to the source? This is a much harder question, and answering it consistently demands particular references. A remembrance of the concert hall is not enough. In perceptual psychology, you begin to understand the tunability of the ear, selective memory, and the ear's ability to adapt. And, of course, we all hear somewhat differently. When you have a number of tuning options in a circuit and don't know which one is right (or are concerned that your observer biases are going to intervene), you have to go to a higher level and make a choice that's more objectively based.
We think that, imperfect as it is, the best basis for choice is the live music experience, and particularly what the microphones capture as a feed. We may or may not be able to capture the sound of the concert hall, which is a function of the miking and that sort of thing. But we can, to a great degree, use the source of a recording as a reference for comparing components that are put in the chain. If possible, we want to get our circuits into the recording chain and listen to the contributions or subtractions.
We do that by inserting the circuits during live recording events on an experimental basis or by inserting them between the microphone feed and loudspeakers. Compared to any recording, the microphone feed is so extraordinarily rich, detailed, spatial, and subtle that to develop high-end audio on the basis of a recording is an obscene compromise.
There is no comparison to what is done under live conditions--the energy, the subtlety, the million-to-one ratio in dynamics. This is all possible with a live musical event, but not with playback. Anything that affects the complete dimensional and dynamic package of sound from the microphones should be looked upon suspiciously, even if it sounds better to you. It can take years to develop an audio component this way--constantly taking the circuits out to the site, listening under live conditions, coming back and making the changes, and also using original lacquers and masters as a reference. But it can ultimately lead you to a level of sonic refinement that is not other wise possible.
Let's move on to the introduction of CDs and CD players, an event you've described as a "trauma," and Spectral's response to it.
CD was a mid-fi medium, a commercial medium with tremendous sex appeal and convenience that addressed almost none of the issues of precision and resolution that Spectral has been involved in. We began to realize that the mainstream industry was not going to address the issues of digital artifacts and digital compromise because of CD's widespread acceptance. In those days, there was no one in the mainstream who said there was anything wrong with digital. We had to roll up our sleeves and get started, and at that point-shortly after the first Compact Disc recordings and players were introduced-there weren't even names for the digital distortions. We lost no time in putting our engineering resources to work on a high-end digital playback system. The result of that work, which Keith Johnson headed, was the SDR-1000, which probably lays claim to being the first from-the-ground-up American Compact Disc player. Previously, CD players were modified with output sections or various tweaks. This was the first time that high-performance discrete circuitry, balanced decoding, conjugate filters in the digital and analog domains, and direct-clocked conversion were used.
The SDR-1000 was introduced nearly a decade ago. More recently, Keith was involved in devising High Definition Compatible Digital encoding. Do you want to comment on what HDCD does for Compact Disc reproduction?
Pacific Microsonics [the company that developed the system, where Keith Johnson is also chief engineer] can make its own statement about what HDCD does, but the antecedents, without question, can be found in Spectral's original digital products. That is the research foundation for the HDCD system that Keith later developed.
Adapted from 1998 Audio magazine article.
Also see:
The Audio Interview: Rudy Van Gelder--Capturing the State of the Art (Nov. 1995)