AO/Beck: Hi Mead. I think you are arguably one of the most colorful and interesting people in the profession. I appreciate your speaking with me today.
Killion: Thanks Doug, it's a pleasure to be here.
AO/Beck: Mead, I think lots of people in the industry and the profession know you from Etymotic Research and the K-AMP circuit, but before we get into technology and related issues, let's start with your education and your early career, if you'll share some of that with me?
Killion: I received my bachelor's degree in mathematics with an economics minor from Wabash College in Indiana in 1961. I worked my way through college doing TV servicing and anything electrical that anybody would pay me for!
And then a little while after college I took a job at Knowles Electronics near Chicago.
AO/Beck: What was your position at Knowles?
Killion: I started as a junior engineer and graduated to designing microphones and receivers, earphones, and instrumentation.
AO/Beck: How long were you with Knowles?
Killion: 21 years.
AO/Beck: What was it that inspired you to go back to school? After all, you had a bachelor's degree and were working full-time at a job that you apparently liked?
Killion: Yes, but I didn't have an engineering degree and I had agreed to take electronic engineering courses after I arrived. As it turned out, I took one engineering course and then continued in mathematics.
AO/Beck: Anything to avoid engineering courses, huh?
Killion: Yes, well, maybe that was it. I actually had a superb teacher in Elmer Carlson, and about five years after I started, I completed coursework for my masters, but I made the mistake of going to a school where you had to have a master's thesis. One of the questions Dr. Menger gave me was fascinating, and it addressed two-dimensional surfaces in Euclidian four-space. I couldn't solve it. I worked on it for about five years and then I finally took my two weeks vacation and locked myself in the workroom at home. By the end of the two weeks, I had the essence of a solution and I was able to finish it up as a combined masters and Ph.D. thesis. So I was pleased to get my master's, but in order to get my doctorate, I realized it would take another few years for the required coursework and I decided I liked engineering more than pure math!
AO/Beck: Nonetheless, I know you earned your Ph.D. somewhere along the line.what happened next?
Killion: You're right. About five years later it was time to start working on the next degree and I went back to Northwestern where I collaborated with Tom Tillman.
AO/Beck: I know his name from Carhart -Tillman.
Killion: Correct. He became my advisor for my PhD in audiology. My research topic was the design and evaluation of high fidelity hearing aids. It was a wonderful education and it tied together many things I was interested in. Eventually I graduated with my Ph.D. in 1979.
AO/Beck: And you were still at Knowles at that time?
Killion: That's right. I was a part-time student for many years, and then I stayed with Knowles several years after I received my Ph.D.
AO/Beck: And then in 1983 you left Knowles and founded Etymotic Research? And I should point out that the name is pronounced et-im-oh-tik, not eat-a-motic.
Killion: Yes, thanks; Etymotic means true to the ear. I left Knowles Electronics twenty years ago. Hugh Knowles was very generous and gave me the rights to three projects. That was unusual because I was under a strict patent agreement when I initiated those ideas; legally they belonged to Knowles. Nonetheless, they allowed me to pursue those three ideas in my new venture.
AO/Beck: Which three ideas were they?
Killion: First there was the insert earphone, then, there was a high-fidelity hearing aid circuit I had designed for my Ph.D. research, and finally there was an infant hearing screener. In addition, I hoped to finish the development of the Class D amplifier that I had been working on at Knowles since 1964.
AO/Beck: OK, let's go one at a time. How did the idea for the insert earphone come about?
Killion: I was taking an audiology course and looking at the World War II headphones. They vibrated the head so much that with a 50 dB conductive loss you didn't know which side was being stimulated. It seemed there should be a better way of delivering sound to the ear. Our first model was the ER-1, with a perfect high fidelity response, and no one wanted it. We sold a couple of them for use in research. Then we made an ER-2 with an eardrum-pressure response that was flat out to16 kHz. We intended it for brainstem testing, but after it failed to give the normal latency-intensity function Wayne Olsen explained to us that it was precisely the lack of bandwidth in the TDH-39 that produced the normal latency-intensity function. Like the ER-1, the ER-2 sales have been mostly for use in research. The third insert earphone, the ER-3 introduced in 1984, was finally a real success.
AO/Beck: And I can remember all of a sudden the world changed when that was introduced.
Killion: Yes, it did solve a few problems. One of the problems it solved was to keep us in business because we were about to run out of funds at that time!
AO/Beck: I guess the circuit you mentioned earlier eventually became known as the K-AMP?
Killion: Yes, that's correct. It was made possible by two SBIR/NIH grants and ultimately it did very well. We've sold three million of them over the years.
AO/Beck: That's really amazing. What about the screener? What can you tell me about that?
Killion: The original infant screener was going to use a cassette-recorded version of an ABR click. We made it for Jack Clemis, a Chicago ENT physician. It was really his idea. The idea was to tape record and present the ABR clicks using one channel, and record the response on the other channel. Then you'd send the tape recording in to somebody with a high speed playback computer that would run it at 10 times real time, so in five or ten minutes you would have the hour's recording and make a report.
AO/Beck: That's amazing. Very simple and it seems to make sense. So, in 1983, you opened Etymotic with these three ideas and the hope you could finish the Class D amplifier. How many people did you have?
Killion: Not many! Ed Devilbiss was a friend who'd done some consulting with us and he said, I'd like to be part of starting a business. He had his MBA in marketing and he also kept the cash-flow spreadsheets at the beginning. He really made the difference in our success. We hired a wonderful school librarian who took orders and kept the books. After about a year we hired a young student. He's still with us and is in charge of our production, a great engineer. We asked if he knew a friend of equally high integrity, so he suggested our second engineer, who was working in a clothing store. We hired him too. They both went through the bad times with us when it looked like we were going to have to shut the doors -- before we got our first ER-3 order. So these guys have been with us a long time.
AO/Beck: Okay let's fast forward to the current product line? What are the different products you're involved with at Etymotic?
Killion: We have a wide variety. I'll start with instrumentation. We have insert earphones for audiometry, ABR and research, probe microphones for Real-ear measurements and low-noise microphones for OAEs. We even have an earphone which comes with 20 feet of silicone tubing so you can send pleasant sound to people while they are in an MRI unit. They can listen to music or whatever it is you want to pass to them.
AO/Beck: Instead of listening to the thump, thump, thump. That's very clever.
Killion: Thanks. Then also in the category of instrumentation, we have the occlusion effect meter. It's a combined seal meter and occlusion effect meter. It allows you to use the subject's voice to check the occlusion effect and the clinician's voice to check the seal.
AO/Beck: So category one is instrumentation. What's the second category?
Killion: The second category would be insert earphones for personal listening and stage monitoring, and hearing protection products such as Musicians Earplugs and ER-20 high-fidelity earplugs that are one size fits most. Our reference-quality ER-4 earphones are used by audiophiles and recording engineer, and our new ER-6 earphones are now offered through Sky Mall.
AO/Beck: Can you tell me a little bit about the ER-4 and ER-6?
Killion: Sure. The frequency response of both is very similar to that of the ER-1 from ten years before. Both the ER-4 and ER-6 attenuate outside sounds and mimic the open-ear response so the experience is as close to live music as possible. The ER-4 provides somewhat higher isolation and a slightly more accurate frequency response. We recently found a partner in Japan that had a manufacturing facility in China, which resulted in the lower-cost ER-6 model.
AO/Beck: And these are not custom made are they?
Killion: No, they're one size fits most.
AO/Beck: So we've got category one, which was instrumentation, and category two was hearing products. What's category three?
Killion: Hearing aids.
AO/Beck: Etymotic doesn't actually manufacture hearing aids, but you do manufacture hearing aid components right?
Killion: That's right, we have hearing aid chips and hybrids, and directional microphones. The K-AMP is pretty well known, I believe. To our knowledge, our sD-MIC and CC-MIC directional microphone designs produce the highest directivity for ITE and mini-canal aids, respectively, of any available first order hearing aid directional microphones.
AO/Beck: What do you think the maximum possible improvement in signal-to-noise is?
Killion: That's a difficult question because different people report different measurements. Measured at the angle of maximum rejection in an anechoic chamber, you can obtain 20 dB, but in the real world where noise from behind is reflected from the front, only with very careful design or multiple-element array microphones can you exceed 5 dB. We measured 17 different directional hearing aids recently, and most of them showed AI-DIs in the 2 to 4 dB range. In fact, we have a patent on anything over 4.7 dB because we were the first ones to show how to achieve that result. At the time that was designed, many people were getting only 2 or 3 dB AI-DI. . Our Link-it array microphone combines the outputs of three directional microphones and has an articulation-index-weighted directivity index (AI-DI) of 7 dB on KEMAR.
AO/Beck: Is there a theoretical ceiling as far as the maximum improvement in signal-to-noise ratio?
Killion: On a single cartridge, a directivity index of 6 dB is the theoretical ceiling measured in free space. On KEMAR or the real ear, it's hard to get much above 5 dB based on the experiments of ours and others.
AO/Beck: I gather you've spent a lot of time and money on hearing aid experiments?
Killion: Yes. We spent about $500,000 a year for seven years optimizing directional microphone designs for hearing aid and automotive use. I once jokingly estimated that it cost $100,000 for each 0.1 dB improvement in AI-DI we achieved.
AO/Beck: Am I correct that for every dB of signal-to-noise improvement, you're going to get a word recognition score improvement of some 14 to 18% or so, is that about right?
Killion: That would be true for a test that used spondees or complete sentences. If you use words in sentences, as we do in the QuickSIN Test, the slope for normals is about 11% per dB and for hearing-impaired subjects it averages 7% per dB. So you might expect 7 to 10 percentage points improvement for every dB of noise reduction using such tests.
AO/Beck: Very good. And I suppose that we can mention amplifiers here in category three too?
Killion: Yes, that works. We make amplifiers too. The K-AMP was the first one and Digi-K is the most recent.
AO/Beck: What year did the K-AMP come out?
AO/Beck: And for those who are not initiated, if they look at that circuit under a microscope they'll see a handlebar mustache right?
Killion: That's true.
AO/Beck: Mead, would you like to tell that story, or do you want to just leave that? I know I tried playing my Beatles records backwards, and you can bet I certainly looked for the moustache the first time I had a dog chew up a K-AMP.
Killion: The moustache is indeed there. There are also the initials of the ten people who helped make that chip a reality. And when you're doing trademarking, it helps to have something unique. But really, the main reason we etched that onto the circuit was for fun.
AO/Beck: And I guess after K-AMP came the Digi-K?
Killion: Yes. The Dig-K came out in 2002, and it's the first digital hearing aid circuit with a 16 kHz bandwidth. The sound quality is really wonderful. We've done a lot of listening tests and the sound quality of the Digi-K and the K-AMP are judged by both normal-hearing and hearing-impaired subjects as substantially higher than the average digital hearing aids. You can tell for yourself by just listening to them. The big advantage of digital processing is that it allows us to use electronic damping, something we patented years ago, so we get rid of the peaks in the response by simply tuning them out and then reinsert the proper shape to the frequency response. So it's the first practical means of providing a truly high fidelity 16 kHz frequency response while still maintaining headroom.
AO/Beck: Can you explain the sound demonstrations below?
Killion: Yes, they show that none of the digital hearing aids, including the Digi-K, solve the problem of hearing in noise, but directional and array microphones can make a major improvement.
BRAND 1 Digital Hearing Aid, Noise Reduction ON
BRAND 2 Digital Hearing Aid, Noise Reduction ON
BRAND 3 Digital Hearing Aid, Noise Reduction ON
Digi-K Hearing Aid, no Noise Reduction
Digi-K Hearing Aid with sD-MIC Directional Microphone
Digi-K Hearing Aid with Link-it Array Microphone
AO/Beck: I like being able to listen to examples. Can we turn to what might be called category four?
Killion: Yes, that would be speech tests.
AO/Beck: Which brings me to one of my favorites, which was the Count the Dots audiogram version of the Articulation Index, adapted I guess, by Mueller and Killion?
Killion: That's not really a test, although I think it was a nice contribution. It does help the patient understand what amplification is all about. I was thinking more about the Speech in Noise (SIN) test, and the Quick-SIN which takes only one minute, instead of four. Our most recent one is the BKB-SIN test. These speech tests are useful for hearing aid comparisons and also for aided versus unaided analysis, and can also be used to get a better measure of what cochlear implant patients might be hearing.
AO/Beck: Absolutely. Have you had a lot of experience with array microphones for implant patients?
Killion: Very little. When we've been able to connect them and make them work, we've received some enthusiasm. One patient said it's like hearing with a telescope.
AO/Beck: Wow, that's a nice thing to say!
Killion: But the problem is it's an engineering project each time we get one cochlear implant patient hooked up! This will probably change in the near future as implant processors are designed to incorporate array microphones.
AO/Beck: Thanks Mead. I know your time is rather limited here today and there are still many other topics I'd like to explore with you. Can we meet again in a month or two to explore noise protection and musician's products?
Killion: I would love to; let's plan on it.
AO/Beck: Very good Mead. Thanks so much for your time today, and we'll follow-up soon.
Killion: Thank you too Doug. It's been fun.
Click here to visit the Etymotic Research website.
AO/Beck: Hi Mead. I think you are arguably one of the most colorful and interesting people in the profession. I appreciate your speaking with me today.