Interviews

Atkinson: What was PSB's first commercial loudspeaker?

Barton: I'm embarrassed to say! It was a 12" two-way. [laughter] The company no longer exists, but Canadians who read Stereophile will probably remember a company called Radio Speakers of Canada. The speaker used one of their 12" woofers with a 1" Philips dome tweeter.

Atkinson: When did you start a relationship with Canada's National Research Council?

Barton: The National Research Council of Canada—NRC, as it's called—is a government-funded group of institutions that employ about 1000 Ph.D.s in all of the sciences, coast to coast. The main campus is located in Canada's capital, Ottawa. The NRC doesn't really get too involved in product development or applied research. They do more cerebral research.

In 1974, a fellow called Ian Masters—who now writes for Stereo Review, among other publications—was the editor of Canada's Audio Scene magazine. I had designed a loudspeaker, a two-way 8" with a 1" dome tweeter, which had a comparing circuit and an accelerometer mounted on the cone. Ian called us, and said he was intrigued that a Canadian company was actually taking high-fidelity seriously. At that time, manufacturers in the loudspeaker business in Canada were just building basic rack-system, department-store boxes under Japanese brandnames. So Ian was enthusiastic about finding out what our little company was all about. He told me that I should go and talk to Dr. Floyd Toole at the NRC.

Floyd was involved in the Acoustics Division of the NRC, which is part of the Physics Department. And as I found out very quickly, Floyd was a music lover and an audiophile. I visited Ottawa, he was enthusiastic about my enthusiasm, and I reciprocated. Pretty much the rest is history. I think in those early days, between 1974 and '82, PSB was the only speaker manufacturer using the facility and its anechoic chamber—and its listening room. Because Audio Scene was doing measurements and ended up doing some listening, a room evolved at NRC that is now cited in many articles as the prototype for the IEC-recommended listening room.

Atkinson: When I visited the NRC a few years back, I was surprised that that room was smaller than I had anticipated.

Barton: Yes. But other than having a high ceiling, about 9', it's typical of rooms in normal houses. The NRC room was also treated with damping materials, curtains, shelving, bookshelves, etc.—to try to "neutralize" the room. By that I mean you could put four loudspeakers in the center of the room and sit two to three listeners in the listening area, and the transfer function was very similar. The room doesn't add a lot of positioning characteristics. The tradeoff, of course, is that you end up with a room that has little room gain at low frequencies.

Atkinson: And, I assume, you also benefited from the research Floyd was doing into loudspeaker performance.

Barton: Yes, Floyd generated a huge database of listening experiences. There is [a series of] Journal of the Audio Engineering Society papers that cover that era of the NRC, between 1974 and 1982 (footnote 1). For the sake of simplicity, that research boils down to three things, the first of which was that most of the people, most of the time, agree on the relative qualities of a group of loudspeakers. That was a revelation, because back in those days you had all these audio people who spoke about "east coast sound" and "west coast" sound.

Second, musical appreciation and experience and taste don't seem to be a prerequisite for someone being a good judge of sound. That should be a revelation to the guy who thinks he can't appreciate quality. Most people, when asked to do it, can fairly quickly learn what the ropes are and appreciate what it is they're hearing, even if they've never done it before. The hard part is to give listeners some way to communicate what it is they're hearing. The difference between an experienced listener and one who really hasn't paid any attention to evaluation of sound is how long it takes them to make a judgment. But they tend to come up with the same answer.

Atkinson: If you think about it, unless our hearing has been damaged in some way, we all have the same basic equipment. The thing that distinguishes audiophiles is not that we hear better, but that we care that there's a difference.

Barton: The final thing that you could say came out of Floyd's work at the NRC—and it's still ongoing research: that a properly interpreted set of loudspeaker measurements correlates very strongly with listener preferences.

Atkinson: Out of all the possible measurements for a loudspeaker, what are the important ones? Or is there a hierarchy?

Barton: There is a hierarchy—at least I think there is. Each designer has his priorities. Certainly if you don't have the frequency response in conjunction with the sound power intact, you can't win. The on-axis response and the sound power work in tandem with each other: You can have a speaker with flat sound power but that doesn't have flat frequency response; and you can have a speaker with flat on-axis response that doesn't have anything that resembles equal energy over the 360 degrees of sound power. They're not mutually exclusive, but you can't have both—unless it's an omnidirectional loudspeaker!

Atkinson: There was a fashion in Germany some 20 years ago for traditional moving-coil loudspeakers to be designed to have flat sound power. But if the speaker is directional, which almost all designs are, if you want a flat power response you're going to have to...

Barton: ...beef up the tweeter...

Atkinson: ...which means the speaker is going to shriek at you.