by David Lander
In one sense, Laurie Fincham came along too soon. Had he entered this world just two hours later, his birth elate would have coincided with King George VI's coronation day. Every child born in England on that auspicious occasion was presented with a free baby carriage.
Yet many would agree that Fincham, a long-time KEF engineer who now holds the title of technical director at this Kent, England loudspeaker firm, arrived on the scene at precisely the right time. Not only can he be credited with numerous advances in KEF speakers over the last two decades, but he pioneered the use of computerized measuring techniques in loudspeaker development and production.
Fincham, whose given name, Lawrence, shortens to Laurie in England (he was surprised to find out it's a woman's name in the U.S.), was born in Southampton in 1937. During the war years, he was sent to a private school in Wales, which he remembers fondly. He later studied engineering at the University of Bristol, where he was more interested in playing bass in a jazz band than he was in applying himself to the curriculum.
Once professionally involved in loudspeakers, however, Fincham's thirst for knowledge of his chosen subject grew. He even returned to school at age 28 to learn more about acoustics. Finding once again that formal education wasn't a proper route to the understanding he sought, Laurie began to pave his own avenue to advances in speaker design. In this interview, he describes a few of its significant landmarks.
The author wishes to thank Dick Moore for his insightful help in preparing this article.
--D.L.
Laurie Fincham on string bass, with violist Bill le Sage sitting
in on piano (ca. 1959), at the first anniversary of the Jazz Club
Music and hi-fi were among your interests even when you were a schoolboy, so why don't we begin with some recollections of those days?
I went to grammar school when I was 11. Keen sportsman. Played a lot of cricket and tennis. Minimum amount of work. I was keen on model aircraft, explosives, music, girls--the usual things.
Explosives?
Oh, yes. I was a great bomb maker. I would have made a wonderful anarchist. The beginnings of my engineering came with making my own cannons. I designed them with recoil mechanisms to improve their accuracy. And I used to make delayed-action bombs, remote control bombs--all sorts of things.
A young friend of yours introduced you to hi-fi at around this time. Was he a fellow bomb maker?
The Jazz Club was run by four members of the John West soup, one
of whom was Fincham, shown here playing bass with violist Bill Le Sage, baritone
Deist Ronnie Ross, and pianist Stan Jones.
He was the ringleader. He was a bit eccentric, but he could afford to buy equipment, and he had one of the first Quad electrostatics when he was about 16. That would be about '54, '55.
He was also quite a keen musician, a guitarist, and introduced me to modern jazz. After a while, I got into model aircraft, so there was a sort of choice between becoming an aeronautical engineer or electrical engineer. And as I didn't want to go into the army, I chose electronics, a deferred occupation. So I went to university when I was about 18 to study electrical engineering.
But of course, in those days, there really wasn't any electronics to speak of. We didn't have a single transistor in the department the whole time I was there. They had just one loudspeaker, and they would encourage you to measure interesting things like accession to inertia. It was heavy engineering--hydroelectrics and so forth. So while I was there, I did-again-a minimum amount of work, but I played in a band. I took up the bass when I went to university. We used to play be-bop with berets and dark glasses and little goatee beards.
And I used to act in revues, write comedy scripts, and generally do anything that didn't involve electronics. I didn't have the interest, you see. When you go to school, they get you on this treadmill when you're 15. After about five or six years, I was fed up with it.
What did you do after you graduated?
When I left university, I did an apprenticeship with Re-diffusion, and I played a lot of music. Re-diffusion are an enormous combine. They owned a TV company; they were into music distribution, they used to make flight simulators, r.f. welding machines, endless-tape players for background music. They used to take on students, and you'd spend maybe a month or three months in various departments. So I did work on flight simulators, which in fact was very interesting because they had speakers there where they were trying to simulate taxi rumble, turbine whine, and so forth. The first job I had was at their research department, which was right at the beginnings of stereo. Re-diffusion used to sell these really nasty little speakers that they would put in hotel bedrooms. It was a cable, and they had the idea that they would distribute stereo and sell twice as many speakers and charge more for the service. I became involved in doing some work looking at stereo and played some of the early Capitol tapes with artists like David Rose and Gordon Jenkins. So I did a bit on speakers there, and after that they offered me a job somewhere in the north of England. I was playing so much jazz at the time that I wasn't interested, so I packed up and went more or less full-time as a professional musician. I used to write music all night and play. And then, during the day, I'd sleep.
Did you envision a career as a professional musician?
No, I didn't really. I liked playing, but I wasn't good enough. I wanted to be a really good bass player, but I realized that I was never going to make it.
Are you still friends with any of the people you played with then?
The pianist. We've been friends for over 40 years. Very good player. He has perfect pitch, which is an absolute pain in the ass if you don't have it. He can sit down and play anything, just hearing it first time. For a time, I thought every [musician] could do that.
Mozart could.
Well, that's the point, I suppose. And if you've knocked around with Mozart, I should think he'd probably put you off any thoughts of becoming a musician.
So you got into loudspeaker design at Re-diffusion.
I didn't design speakers. I just got the feeling that speakers might be a good thing to do. Then I saw an advert for Goodmans. They were looking for engineers. So I went along there and got a job as a junior development engineer. And I did what everybody else does who gets into speakers--I did all the wrong things. You get into speakers and you think, "Oh, good idea to make stiff cones." I was playing around with expanded polystyrene, things like that. There was a lot of that going on in the late '50s, early '60s. So I did that for a bit, and then they found out I was a musician. Goodmans were pretty old-fashioned at that time, very much a '30s company, and so they used to get me to deal with all the musicians who called at the door with broken speakers. I ended up designing quite a lot of musical instrument speakers-this would be about '62, I suppose, just the beginning of rock and roll. The big revolution had come about in the late '50s, when I was still a student, but, being a good be-bopper, I looked down my nose at rock and roll. I thought it was terribly beyond the pale.
-------------The Maxim loudspeaker, which was sold in the U.S. under
the name Maximus, was designed by Fincham while he was working as an engineer
for Goodmans. It was his first commercially successful design.
How did rock and roll change things?
I think what happened was that, all of a sudden, the speakers people had been making over the years suddenly became too feeble for use with musical instruments. When I started playing in bands, no place had P.A. There'd be one carbon microphone and a little 10 inch open-baffle speaker stuck up in the corner. Rock changed all that.
They had to have stage P.A., and they had to have amplifiers on stage. In the early days, designers in the U.K. designed everything in a classical, orthodox way, so you had hi-fi amplifier designs being produced for guitar amplifiers, and of course that was totally wrong. Fender over here knew perfectly well how to do it. They just designed amplifiers with very little feedback. So I used to travel around to various guitar manufacturers, just talking to them about how guitar amps should be designed, and then I started on a miniature hi-fi speaker called the Maxim. It was sold over here under the name of Maximus.
Tell us something about it.
In the very early '60s, there was a company called Leak, and there was a man there called Don Barlow, a very interesting and quite innovative engineer. He made a tweeter that I think really was a rip-off of a Wharfedale tweeter, which was a sort of straight sided 3-inch cone with a cloth surround. In the lab at Goodmans, at the time, was a chap called Ted Jordan E. J. Jordan. He'd seen this cone, thought he could make a miniature full range speaker based on the design, with lots of bass, and he did a bit on it.
He'd stopped playing with this thing, and I decided I would develop it into a little full-range drive unit, but it didn't work out. Then Goodmans took an interest in putting this thing on the market. I panicked. We had about six weeks to go, so I designed a little cone tweeter to go with it, and they did extremely well with it. What was funny was that, later on, it got copied quite a lot by other people, and they copied all the things that were wrong! The reason we were using these huge magnets was that somebody in the buying department had over-ordered. They bought 10,000 of these large steel magnets and said, "Use the damn things up." But of course they were too big, so I had to open up the gap. And they were too mean [British for "cheap"] to spend money on new voice-coil formers, so I made a very fancy plate shape so it would reach into the gap. It was amusing to see how people copied it without really understanding it. There wasn't anything fundamental about the thing. It was just fortunate that it was the break you need early on in your career when they say, "Who did that?" So I then got offered a job by Celestion because the former managing director of Good mans had left and gone there.
Tell us about your tenure at Celestion.
They didn't really have a hi-fi department. I think that when I arrived there, we didn't have an oscillator. We had to borrow it from one of the subsidiaries in the group. I think we had an AVO meter, which is a very old-fashioned moving-coil meter, and that was about it. Celestion were an even older company than Goodmans, but they were extremely mean. Whereas Goodmans had an anechoic chamber, lots of equipment, and lots of graduate engineers, Celestion had no anechoic chamber and I was the first graduate they'd recruited in 29 years. They used to call me Einstein and a few things like that. But it was a great place to work because there were very, very good mechanical engineers there. I worked for an old fellow called Les Ward looked like Sir Adrian Boult, actually.
He was a bit of a spiky character. He was a former boxer, I think, and used to play the trumpet in a band. He was a very good mechanical man. He taught me lots and lots of things about mechanical speakers, about horn speaker design. He was totally intuitive, completely self-taught.
"Totally intuitive" sounds like a lot of people's conception of the ideal speaker engineer: Three parts magic, one part science.
People like to think there's witchcraft in speakers. Now I'm not here to say there isn't. A good chef, with the same ingredients, will always cook a better meal. What I am here to say is that an awful lot of it isn't witchcraft. It's just lack of understanding. If we just get rid of the perhaps 90% that is amenable to analysis, then we could get on to the bit that really matters.
In fact, you've pioneered a lot of the analysis that's been done in the speaker industry during the last 15 or 20 years. One of your principal activities since moving to KEF has been computerized measurement. What got you interested in that?
About 1969, I had a call from a man in Bradford University called Rex Leed ham, a genuine eccentric but a very innovative character. He wanted to simulate the acoustics of a living room, which was quite an advanced thing to be looking at in 1969, because there weren't many computers about. He had with him a very bright young student called Mike Berman, who subsequently came and ran our research department. He was doing a Ph.D. on the simulation of room acoustics. Anyway, we got together. At this time, I'd been working with speakers for about eight or nine years, and I realized I knew very little about what I was doing.
It hasn't changed much. I'm bound to say, 20 years on, we're still as dumb as we were then. It's just that we thought we were smarter. So I thought there must be a way of measuring speakers other than with sine waves. I've always had this feeling that we use the wrong signals, the wrong instruments, to test speakers. What we do is use the instruments that are to hand. We don't ask ourselves whether that's appropriate. I think we should test speakers using the signals that the speakers are actually going to be used for--music. A lot of the work I've been doing over the years has been trying to get signals that are closer to musical sounds than they are to sine waves.
You're getting ahead of yourself. Tell us about the first computer experiments you did.
We started measuring the transient behavior of speakers using pulses. We bought an instrument called a single point correlator to do this. So I said to the salesman, "What are we going to do with that?" He said, "Well, can't do much. It's just the pulse response." And I said, "Well, I can't tell anything by looking at that." "Ah," he said, "what you want is a Fourier analyzer.
It'll analyze the frequency content of this pulse." I said, "How much are they?" "Oh," he said, "they're about 14,000 pounds." Now in 1970, 14,000 pounds was like $300,000, so what we did was obtain a grant from the government-the university did-which enabled them to buy one of these Fourier analyzers. Rex Leedham and Mike Berman did a project there and also acted as consultants for KEF. They did this for three or four years, I suppose, and every weekend I used to travel up and down the motorway to Yorkshire, which is about 200 miles from where I lived, to see how the experiments were coming. In the end I said to Mike Berman, "You'd better come and work for us because I'm fed up with doing all of this." So he said he'd come down to Kent, and he did in 1975.
Famous British electronics engineer Peter Baxandall aided
KEF's work on the 104/2 when he suggested that the driver be put into the
box, rather than on its face, and that servo feedback be used.
And KEF bought its own Fourier analyzer at this time?
Before we bought it, I went to HP in California. Now these things were used on nuclear subs and by car manufacturers. They weren't used in the audio industry at all. I said, "We'll buy it on one condition. You've got a mass storage device. Surely we can record music on it." So they set up these experiments, and I think we could sample at 20 kHz, which gave you a maximum frequency of 7 kHz. But the idea I had at that time was this: If we could record music, store it on the computer, and measure the speaker digitally, then maybe we could simulate the effect on the reproduced sound of changing the measured performance. When Mike joined, we developed a scheme for recording directly onto a computer disc and did a presentation, I think in 1975, at the AES in London, where we recorded on a hard disc, a removable one.
How much signal could you store?
We could get 20 seconds of music on it. It was 14-bit resolution and 79-kHz sampling--pretty advanced for those days. The A-to-D converters we used cost us, then, somewhere in the region of $8,000. What we did was record the music on the disc, then digitally filter the music as if it had been played through a speaker. We used a technique called convolution, which is pretty slow. In fact, it used to take about eight hours to pass 20 seconds of recorded music through a digital filter. We couldn't do this during the day because it would hog the computer, so Mike Berman had a bed in there. He used to switch the thing on, go to sleep, and then have to wake up halfway through the night because the computer didn't have a very wide dynamic range. At that point, he'd go through the data looking for the largest value so he'd get good signal-to-noise. Well, as you can imagine, the glamour wore off after a few nights. So we didn't really do any more with the technique. In fact, it's taken from then until now to have real-time DSP machines that can do what we wanted to do 15 years ago. We were frustrated by lack of equipment, so we just used the computers for measuring speakers initially.
You've relied on a number of universities for personnel and research over the years, which is not something American speaker companies do. What's your rationale for this?
What we've found is that you cannot have all the skills that you require within a research department, even when it's pretty well equipped and fairly well staffed. Even with six or eight engineers, you couldn't have all the skills you wanted.
When I want to get somebody to work on a project, I go and find the best man in that area, [the one, for example,] who knows the most about DSP or acoustics. So from our point of view, going and working with a department of a university was a very good thing for us to do. It was never done hands-off, mind you. It was always directed research. We would certainly go along with a list of things we wanted to know and give them some sort of time scale on it. We must have worked with eight or ten universities, I suppose. But I do a lot of lecturing at universities, so we've always had that link with them. In fact, the man who runs our research department, Dick Small, was an academic.
Right, in Australia. How did you come to meet him?
I've known Dick for ages. He was doing a Ph.D. in Sydney, and he'd measured some speakers I'd designed. That all really stemmed from Neville Thiele. I'd got hold of Thiele's famous paper on reflex enclosure design, which he wrote in '60 or '61 when I was at Goodmans. I didn't understand a word of the thing, but I persevered with it and eventually put in measuring system equivalents that used a lot of Thiele's techniques.
Dick Small, when he was doing his thesis, was surprised to find a company that appeared to understand what this was about. Thiele was not as well known as he is now.
In fact, Small first worked for KEF while he was on sabbatical from his university job.When was that?
I think it was '84. He stayed with us six months or a year; the time just flew by. Then he went back to Australia.
above: A live-versus-loudspeaker demonstration featuring, jazz
bassist Red Mitchell in a showdown with the KEN 104/2 speaker system. Such
a test pits the speaker against its true signal--music.
How did you get him on a fulltime basis? A lot of people would agree that was a great coup.
At that time, there were a lot of cutbacks--more students, less time for research--and he wanted to do research.
And the equipment he'd so carefully built up for pulse testing had all been cannibalized for other purposes. I think he became pretty disenchanted with academic life. I went over there, and he voiced that. He said, "I'm really thinking of getting out." So I said, "Give us first refusal." He said, "Well, I didn't think you'd have me." mean, that was Dick. He's a very modest soul.
In addition to computerized measurement, you're well known in the speaker industry for your work in the area of low-frequency response.
That's the easy part. That's the reason [laughs].
Well, let me combine two thoughts here. For one thing, you've said that certain things you've worked on have been developed in stages over a period of many years. The coupled cavity in the KEF 104/2 is an example of that and also represents your and also represents your work in the low-frequency realm. What can you tell us about it?
We were working on a servo system, in the early '70s, with Peter Baxandall. Baxandall had this idea, and he said, "If we just put the speaker inside the box and listened to what came out of the port and used servo feedback, we could make a really good bass speaker." I got to thinking about that and thought, "Yes, but we could do that with a passive system." I thought it was new at the time, and I did an analysis over Christmas in 1974. Within two months, I got a letter from a good friend of mine named Siegfried Linkwitz, who worked for Hewlett-Packard. He said, "Look at this thing by Onkyo. It looks like the thing you've just described." So I thought, "Forget it; it's not even original." And we left it on the shelf for about five years. Then I thought, "Well, it's interesting. I'll write a paper on it." So I did a paper in '79, and then we produced the product in '84. It had been hanging around for 10 years. It was on the shelf, and I thought, "Well damn it, I'll make it work." And we did, finally.
Then didn't a French company get in touch with you, claiming the idea was theirs?
Elipson. They said they'd invented it and that they'd patented it in 1968. I was sure they didn't have the exclusive patent on this thing, so I went on a great patent search. We then heard through our Danish agent--this is how these things go--that he had read an article reprinted from a French underground magazine saying that the original design had been done by someone in the United States. Now [KEF founder] Raymond Cooke--who loves a detective tale, just loves sleuthing after things--came over to the States and [went through the archives until he] found a reference to this speaker. It was designed by a couple of people called Baruch and Lang in about 1951. They were the ones who had done all the original work.
above: Basic psychoacoustic research is being carried out in the Archimedes project, a joint venture between the Technical University of Denmark, B & O, and KEF, to show how a room interacts with a speaker.
You're now two years into an extremely ambitious project with Bang & Olufsen that involves the interaction of speakers and room surfaces.
What we basically did was to say that the differences in sound between nominally identical speakers must have to do with the reflected sound in the room--that the way sound was reflecting off the walls was influencing what we heard. That must be due to two things: The directional characteristics of the speaker (the way in which it spreads sound around the room) and the acoustics of the room (size, shape, where you're sitting, and so forth). Now there are so many variables with a speaker in a room that we thought it would be nice if we could deal with one thing at a time. We decided to do it by a simulation method, where we set up a number of speakers, a number of sources, in a completely free space, with each source representing one reflection in a room. So--in theory, at least--we could change the nature of these reflections by just attenuating or filtering the input of the various speakers. We knew from Mike Berman's work in the early '70s that you could analyze a room in this way, make computer models where you used an analogy between the behavior of sound in a room and [the behavior of light if] the walls in the room were mirrors. So we set up a program where we were going to make sound sources as good as we knew how to make them. We would hang them up in this chamber in an array predicted by a computer,[and that] would simulate a listener in a particular room, listening to a particular speaker.
And how far along are you?
That's pretty well where we've got to now. We've got the thing set up in the anechoic chamber, and we'll be able to simulate the effect of changing the acoustics of the room. I could actually move the room, with respect to the two of us, just by flicking a switch. We can change the directional characteristics of the speaker--at least we can simulate the effect of that change. By doing this, we hope to come up with the degree of sensitivity that the ear has to changes in the room characteristics or in the [loudspeaker's] directional characteristics.
So you'll be exposing live listeners to computer-programmed changes in the reflection characteristics of simulated rooms?
Yes, that's right. You sit in a chair in this anechoic chamber and just look into space, because you can't see any of this at all. You're in a net curtain tube, and all these speakers are hung at various positions around you. You have a keyboard, you're given a sound, and you have to key in a response. The psychoacoustic tests are set up to let you hear two things in relation to one another. For example, we could ask, "How important is the reflection from the floor; does that make a difference?" We could program in a floor that was concrete or a floor with a foam-rubber mattress on it or something like that. Now what we're interested in [determining] is, at what point do you hear the difference? We're interested in thresholds.
So you've come full circle. You're back to flight simulators, except that this simulator is for the ears exclusively.
Absolutely.
Will these experiments lead us to the perfect speaker?
Our object in these experiments is not a reproduction system that is perfect. It's a system that is as good as the best available today, no matter where you use it. Wouldn't it be nice if you could make a speaker that always sounded its best, no matter where it was?
But just how complicated would such a speaker have to be? A number of hi-fi designers are now turning to. very complex signal processing. What ever happened to the old audiophile ideal of a straight wire with gain?
I don't like having a lot of technology if you can possibly help it, and I don't expect that the systems that come out of these researches will be that complicated. People have had a go at these room equalizers, and they came up with immensely complicated systems that tried to recreate the perfect signal at the ear. That's an absolute waste of time. You have to produce the right signal at the ear. The ear's very tolerant of the wrong signal in many cases and, [in others], very intolerant of something that's just slightly off.
I'd like to go back to your earlier comment about everything in speakers having been done, or at least having been conceived, before. In the late '50s, Ed Villchur's acoustic suspension was held invalid after he sued Electro Voice for infringement. Apparently, Harry Olson had written about it.
My feeling about patents in our business is. no matter what you think of, it's always been done before. In a way, I suppose. I cling to the view that it doesn't matter a damn who was the first to think about something. It's the first person to do something about it. Go and make it into a product. Otherwise, it's pointless.
It's interesting that loudspeaker technology has changed a lot less than some related audio technologies. The idea of a dynamic, moving-coil loudspeaker, for example, is rooted in another age entirely.
The original moving coil was proposed in 1877. We're still light-years away from other industries. What appalls me about loudspeakers, as I always say to people, is this is an agricultural business.
Australian Richard Small was able to extend his countryman
Neville Thiele's work on box design and driver interaction so that designing
a system is no longer a black art.
Why? Is it that you and your colleagues are slow, or that your predecessors were so visionary?
I think it's a bit of both, really. I think that the brightest people were in audio when audio was an emerging business. All the bright people worked in Bell Telephone, Westrex, and they were all working in the '30s. It was the cinema which was pushing it forward, the need to have high power output and so forth. Then they got into the war, and it was underwater speakers and hydrophones for sonar. After the war, it was TV, so all the people went to TV. Now they're into computers. So for a long time, we did not attract the best people into the industry; it was a peripheral industry. The other reason, I think, is that the moving-coil loudspeaker was actually quite a good idea. It's like the piston engine. Why do we not have turbines or electric motors in cars now? Because [the piston engine] was really pretty good. The trouble with speakers is they work better than they ought to for such a crude device that's relatively inexpensive to make. That's one of the reasons why they won't change. The other point about speakers is there isn't really an economy. Take a TV set [from] 20 years ago: [Today,] you can make it for a fraction of the price. A speaker you can't quite [do that with], because it's a chunk of wood. It costs more money than it did before. All the parts. There's no benefit of miniaturization. Really, when you think about it, there's an indictment. We're not really doing it the right way. All we've got to do is shift a bit of air.
Will it change?
It will change. The thing that will happen is digital processing will go into speakers. There will be other technologies that allow us to reduce the size of enclosures. It will happen. It's happening now.
Can you give us a more detailed picture of this?
Yes. I think what will happen is the front-end of audio will disappear, and the loudspeaker will be the last thing left over-no matter what. I don't think speakers need to get that much better, but they need to be more convenient. I think speakers, in the end, will be things that are plugged in around the house, like table lamps.
But there will be loudspeakers.
You've still got to get something to actually shift the air. It's the last interface with your ear.
(adapted from Audio magazine, Jun. 1990)
Also see:
The Audio Interview: John Charles Cox--A Matter of Precedence (Jan. 1985)
The Certified Bass for the Certifiable (Jan. 1990)
The Audio Interview: Direct Reflections from Amar Bose (July 1983)
= = = =