1. Was that approach something you had to request, or did Sidney Harman bring up the idea of continuing to publish?
2. One of the things that Sidney Harman was a big advocate of was having research, design and manufacturing in a single location for the fastest transfer of ideas to product and to allow cross pollination between the groups.
Do you have any stories that you can share of how something went from a discovery in your lab to a product?
3. One particular question I had was when Revel went from rear firing tweeters in the first generation speakers to a more traditional setup in the current generation products. Are there certain rooms where the added energy is helpful?
4. Are there any university groups still involved in this field of study? Or as home audio market has dwindled has academic interest also slowed down?
5. Given the ready availability of tools like the Klippel to analyze speakers without an expensive anechoic chamber, and the fact that a handful of billionaires are audiophiles, what sort of endowment would it take to actually set up an lab to continue some of your work in this space? Asr is Amir‘s retirement hobby but imagine a 501c3. (Is this a $10M or a $100M effort?)
6. Last, any comments on the JBL Paragon? Both for its sound in its era and today, if the constraint was furniture of similar proportions (i.e. a single speaker instead of two separate towers).
Thanks for all of your input in this forum.
"Was that approach something you had to request, or did Sidney Harman bring up the idea of continuing to publish?"
As a research scientist I was a bit of an anomaly in the context of this large manufacturing company, many people thought I was an academic, detached from their world. In fact, at the National Research Council of Canada (not a university!) my research had led me into intimate interactions with several loudspeaker designers and manufacturers - for about 20 years by then - all very much real world. They were benefitting from the guidance of my research, renting the NRC anechoic chamber and custom measurement equipment to design their products, and using the double-blind listening test to verify their efforts. It was the beginning of the Canadian loudspeaker industry which I am proud to see is still prospering.
I was hired because of what I knew, and that came about as a result of applying the scientific method to investigations of the unknown. Doing it required engineering facilities that were lacking at Harman and subjective evaluations that were more than the sighted opinions of the designer and a few management-level people. "More bass, more treble" was not an uncommon parting statement from marketing types. It was insulting to some serious efforts by engineers, and it is easy to see why loudspeakers in the marketplace varied so enormously in sound quality.
Nobody could remember a blind, much less a double-blind listening test ever having been done at Harman. "We are professionals" was the reason they could not be biased. Sean and I managed to get several of them to participate in a test of their objectivity:
Toole, F. E. and Olive, S.E. (
1994). “Hearing is believing vs. believing is hearing: blind vs. sighted listening tests and other interesting things”. 97th Convention, Audio Eng. Soc., Preprint 3894. It was a wake-up moment for all but one engineer, who persisted in avoiding having his standing as a "golden ear" tested. I was a corporate officer and had no line authority over engineers employed by the brands. I could only suggest, and to the brands what mattered most was sales.
In discussions in this forum it is clear that some participants just don't realize what a rudimentary state the loudspeaker design was in back in the 1970s and 80s. Measurements were primitive. One of the most prominent high-end loudspeaker companies of the day had only a 1/3-octave real-time analyzer to do room curves - an approximation to no useful data at all. Real decisions were made in informal listening sessions, all sighted - all possible subjective biases in place - and some while "under the influence" (smile). I felt I had created an oasis of calm and sanity, where if the objective is to design a good sounding, neutral timbre, loudspeaker it was possible to do so with confidence - as well as to deviate knowingly from that if it is felt that the market desired it.
Soon there were discussions about continuing the research at Harman so that we could refine what was known and expand new knowledge into areas that were beneficial to our products. I asked for, and got, permission to set up a Corporate Research Group to create new knowledge - not new products. A key component to research is the public sharing - a give and take - of knowledge; presenting and publishing scientific papers. Remarkably, it was not a big issue, and Sidney Harman agreed that we could publish. I sensed he felt that it would add to the image of the corporation, separating us from our competitors. He was right, but not in the domain of consumer audio which was improving, but remained significantly confused by "incomplete information" about products, with a skepticism that science and measurements had anything to do with music, soundstage and imaging, air, tight bass and "resolution". It was totally different in the area of automotive audio. This was a rapidly growing sector of our business (about 10% of $0.5M when I joined, and about 75% of about $7B including infomatics when I retired). Automobile engineers respect science, knowledge and facts - they listened intently at my presentations to a "who's who" of the automobile industry over the years.
I immediately hired Sean Olive, my colleague at the NRC in Canada, and a bit later Todd Welti, an educated and experienced acoustician. Sidney authorized millions of dollars of investment in engineering facilities, anechoic chambers (2 pi for transducers and 4 pi for systems) all set up for high resolution measurements and calibrated to low frequencies. Up to date computers and modeling software, a scanning laser interferometer, transducer parameter and distortion measuring apparatus all added insight and capability. New listening rooms were created for double-blind testing, a unique positional-substitution one being the most useful. The listening room cannot be eliminated but it can be rendered a constant factor in comparison tests.
Over the years several smart engineering interns went through the research group and ended up in prominent engineering jobs in Harman and with our competitors. It is a "life changing" experience to witness how reliable and strong the connection is between measurements and double-blind listening tests. Allan Devantier, another Canadian from the NRC period, had a successful career at Harman, being responsible for some of the best designs by Infinity, and co-creator (with Charles Sprinkle, another believer in science, now with Kali) of the JBL Pro M2, a benchmark design. After a brief stay in my research group where he collaborated with Todd Welti on an active multi-subwoofer solution called Sound Field Management, Allan now heads up Samsung's audio engineering activity, and has an updated version of all of the Harman measurement and evaluation capabilities. Samsung later purchased Harman, mainly, I think, because of the access it provides to the automotive industry. Harman operates substantially independently, though - as least as I understand things. I retired in 2007, and remained on a consulting retainer until 2019.
"3. One particular question I had was when Revel went from rear firing tweeters in the first generation speakers to a more traditional setup in the current generation products. Are there certain rooms where the added energy is helpful?"
It was conceived, I believe, as a product differentiator with a rational reason to exist. However, in real rooms its benefits were not clear. One reason is that air attenuation sets in above about 1 kHz and is substantial with increasing distances (Section 10.6 in the 3rd edition). Just do some ray tracing in normal rooms and it is clear that any sound that has reflected once or twice has not only undergone some loss at the reflecting boundary but has lost energy to air attenuation. Result the spectrum of the reflected sound field rolls off at high frequencies. This rear-mounted tweeter had little chance of changing that - but in principal it is logical if measured power response is a consideration.
"4. Are there any university groups still involved in this field of study? Or as home audio market has dwindled has academic interest also slowed down?"
In general universities lack the necessary measurement facilities, listening facilities, and budgets to prepare for serious efforts in this area. As can be seen in discussions in this and other audio forums, many people have moved on from pure stereo, enjoying the benefits of processed (upmixed) or real multichannel movies, music videos and music - always with the opportunity to revert to stereo. Those who by choice or circumstances remain in the stereo domain, still seek something that the two-channel format cannot deliver, and endless debates ensue.
Audio has evolved, and decent sounding baby bluetooth loudspeakers responding to voice commands can deliver tolerable tunes in every room in a house. Headphones abound, and even I enjoy listening to music or podcasts while walking or flying. Sean Olive - now Dr. Sean Olive - continues his research but the emphasis is now on headphones. His contributions are profound, again the result of patient, time and effort consuming blind tests and valid measurements, all published for the world to absorb.
Yes there are a few university efforts ,mostly as I observe, in England, Europe and Canada, and largely engaged in multichannel applications. When I last visited McGill University in Montreal, one of the contributors, all of the post graduate projects that were demonstrated to me involved multiple - more than 5 or 7 - channels. Stereo was very much a "been there done that" issue. But as I see it, the big questions about loudspeakers have substantially been answered, and money to support the motivations to explore second-level and lower issues, especially those applicable to purist stereo, is not abundant.
"5. Given the ready availability of tools like the Klippel to analyze speakers without an expensive anechoic chamber, and the fact that a handful of billionaires are audiophiles, what sort of endowment would it take to actually set up an lab to continue some of your work in this space? Asr is Amir‘s retirement hobby but imagine a 501c3. (Is this a $10M or a $100M effort?)"
A nice fantasy. A few years before I left the NRC I arranged a non-profit collaboration between a consortium of Canadian audio manufacturers and the NRC. It was necessitated by budget cuts at the NRC, a federal government organization, and science was losing favor among politicians (sound familiar?). It was the first collaboration between industry and government research in Canada at the time. Klippel (who was employed for a period in my research group BTW) measurement systems are simply minor miracles from my perspective - real hard science at work. Nevertheless, anechoic chambers still have advantages: a quiet background against which to evaluate non-linear distortions being one, but also it can be a permanent setup into which very different loudspeakers can be introduced and measured with minimal fuss But to answer your question, it depends on the cost of the real estate, its location, the ability to find competent researchers and their cost, the measurement equipment, the subjective evaluation venues and apparatus. Setup: less than $10M - pocket change for the 1%, but would it provide the necessary gratification?
"6. Last, any comments on the JBL Paragon? Both for its sound in its era and today, if the constraint was furniture of similar proportions (i.e. a single speaker instead of two separate towers)."
The JBL Paragon was a simply beautiful piece of furniture - impressive if one is into art-deco/Scandinavian style decor as I am. I have a story. When I was a PhD student at Imperial College in London, I tried to continue my audio hobby interest, but had no budget and no place to set up a system. I followed developments and went to an audio show at the Russell Hotel - tiny rooms typical of British hotels of the era. I had no pretensions about my ability to judge sounds, I could only say what I liked, or not. I went in with some strong expectations, such as: Quad ESL should be superb - "massless diaphragm" eh?, Tannoy, with its concentric design should reflect good engineering - point source, eh?, and so on. Reality set in quickly, and I discovered that simplistic views and marketing claims are a different thing. The Quad indeed sounded good, the ticks and pops of the LPs were reproduced with finesse, and the music sounded pleasing, but it seemed to have problems playing loud when people asked for the volume to be turned up beyond "polite". The Tannoy sounded horrible, shrill, penetrating, ugly. I was crestfallen. So I embarked on a wide survey of what was on display. One room had a JBL Paragon. It went almost from wall to wall in the tiny room. and was elevated to near ear level because of the listening distance. It could play indecently loud, crystal clear, impressive indeed. It seemed to have a distinctive sonic "personality" which was not totally endearing, but there was more to it than that. I returned a few times to try to pin down what that demo had that others lacked. When I poked around I opened a closet door and lo and behold there was a monster half-inch two track 15 inch-per-second Ampex tape recorder. That was it! Master-quality program material! Everyone else was using LPs with all of the inherent distortions, noise, frequency response inconsistencies and dynamic limitations.
So what did I find attractive? The KEF Concord, a new product from a new company seemed to have something going for it. I got to know Raymond Cooke, the owner/desingner, and wangled a price reduction "for research" - tongue in cheek. As explained in Section 18.1 in the 3rd edition with measurements, I wasn't completely wrong, and the speakers never made it into my home as purchased. Raymond and I remained friends, visiting back and forth for several years. Through him I subsequently got to know Laurie Fincham then head of engineering at KEF, who remains a friend here in California, having been hired into Harman via Infinity, then moved to THX. I also got to spend quality time with Peter Walker of Quad, extensively touring the facility, sitting in "his" chair where he "voiced" his loudspeakers, and drinking good Scotch with him in his large living room with a pair of ESL-63s providing the tunes. He was a fine person, nice, unpretentious, and very capable.
In its day the Paragon might have been sonically competitive, but it was far from neutral. One sat in the Harman corporate office while I was there - silent, elegant, a lovely piece of audio sculpture.