The "mad scientist" of Sony audio

When we think about how electronics products are developed, we might imagine huge teams of faceless engineers, executives embroiled in endless discussions in elaborate conference rooms, and an almost Kafkaesque process that no one person really understands or controls. Having visited the overseas facilities of many electronics manufacturers, I can attest that there’s much truth in this conception. That’s why I was so surprised to find, deep in the engineering labs of Sony’s R&D building in Tokyo’s Osaki neighborhood, one man who for many years has guided the development of the company’s home audio products

Takashi Kanai has been at Sony since his career began in 1978. The title on his card—Sony MVP2005, Chief Distinguished Engineer Graduate, Acoustic Design Department, V&W Business Division, Home Entertainment & Sound Business Group—hints at his importance even as it obscures what he really does. From his personal laboratory, filled with stacks of audio gear, bins full of parts, shelf after shelf of test gear, and a huge B&W Matrix 801 surround-sound system similar to those often used in mastering labs, he directs the development of Sony’s home audio electronics, with a particular focus on that most complex of audio products, the A/V receiver. He also maintains his own website where he discusses audio engineering and technology.

In this lab, I interviewed Mr. Kanai to explore his philosophies about audio engineering and learn about the challenges facing the designers of today’s audio products. During our discussion, Mr. Kanai went back and forth between Japanese and English; fortunately, one of Sony’s Japanese product managers was on hand to translate when necessary. I’ve edited the conversation for easier reading, and apologize in advance if I’ve distorted anything Mr. Kanai said.

Visiting an engineer in his lab might seem intimidating, because most are reluctant and unskilled communicators. But Mr. Kanai’s engaging personality, casual demeanor, and passion for his field made me feel right at home, like I was talking with a fellow audio enthusiast rather than one of the most influential engineers at one of the world’s largest consumer electronics companies. As I left the building later, I saw him again in passing, headed for the train with a large pack strapped to his back and a comfy sweatshirt wrapped around his shoulders, looking more like a 20-something New Yorker than an engineer with three and a half decades of experience.

Sound+Vision: When it comes to sound quality, what’s the most important part of an A/V receiver?
Takashi Kanai: That’s a very difficult question, because sound quality comes from many elements—the parts, the circuit, the chassis. One of the main things, though, is noise. There’s so much noise from digital circuits, from the DC-to-DC converters, from the AC-to-DC converter. And now every product has a microcontroller. They all make terrible noise. Our challenge is how to escape from that noise.

What steps do you take to eliminate the effects of the noise?
Much of it is in the design of the chassis. For example, a DSP [digital signal processor] on one side of the receiver can make noise, and if you don’t design the chassis carefully, the circuits on the other side of the receiver can act as receiving antennas and pick up that noise.
Another big problem is vibration. Vibration can change the conductivity of different parts of a circuit. It can also change the values of the capacitors, because the paper and foil used inside the capacitors are weak materials. With digital signals, it doesn’t change the frequency, but the height of the waveform can change. As the height of the waveform changes, it can change the place where the signal breaks the threshold [differentiating a digital “1” from a digital “0”]. And then you get jitter, which degrades the sound. You can see the effects of vibration with an oscilloscope. Hook up the oscilloscope to one of the audio circuits, then tap on the chassis, and you can watch how the audio signal changes.

Receivers are far more complex now than they were 20 years ago. How has that complicated your job?
Not much. Of course, there are several difficulties when you add new technologies, but it’s all fun for me. Except once I felt like I was going to die.

What happened?
In 2002, I was working on the S-Master Pro digital amplifier. At the time, no one knew how to make good sound with a digital amplifier. Sony’s the kind of company where if you’re having difficulty with a problem there’s someone who knows how to solve it. But with the digital amp there was no one who could help me. It got so bad I physically fell down and the ambulance had to come for me.

But you eventually solved the problem?
Oh, yes, that product got the gold prize that year from HiVi magazine in Japan.

Every year, you have to incorporate new technologies into Sony’s A/V receivers, but you have to sell them for the same price as last year. How can you do this and maintain the sound quality?
That question always comes up. When we add technologies like Bluetooth and MHL, it’s fun and simple for the customer, but it’s hard for us because these things tend to add noise. It’s a step-by-step process to learn how to preserve and even improve the sound quality as you add more features to the receiver.

So the improvement in your skills outweighs the negative effects of the new technologies and the cost that they add?
Yes yes yes yes yes! You can’t keep the sound quality unless your skills keep improving. Five years ago, the receivers couldn’t sound as good as the ones we’re introducing this year. Look at this chassis from five years ago [holds up the underside of an older A/V receiver]. There are only a few embossings on the bottom plate to strengthen it. We found that the areas around the feet flex a lot and allow vibration. The area under the transformer also turned out to be weak.
Now look at this new model [holds up the underside of one of Sony’s yet-to-be-announced 2013 models], which we worked on with the mechanical engineer in our Malaysian factory. Every embossing is straight instead of side-to-side, which makes it much stiffer. Even the low-end model from this year is stiffer than the high-end model from last year. We also added vents of different sizes, and altered the spacing of the vents to reduce resonance and damp vibration. None of this has much impact on cost, but it all improves the sound quality.
When you’re building A/V receivers, and you’re pressed to come out with new models every year but to keep prices down, improvement is a gradual process. If I was working at Mark Levinson, I could make this new chassis right away.

Is your work on design and engineering only, or do you get involved with manufacturing also?
Yes, we work with the factories to make sure the product is the best it can be. We even designed our own solder. [Gestures to another engineer to get some solder samples off a shelf.] This is the solder we used to use, made from tin and lead. We used it for 30 years. [Breaks off a piece, holds it in his mouth, and moves the tip toward his fingers holding an imaginary soldering iron.] Of course, we can’t use it now because lead is poisonous. But with lead-free solder, the sound quality is generally very bad. I went through more than 50 samples of lead-free solder, and I found that the tin mined from a certain mountain gave good sound. But then if that mountain closes we cannot make good solder. So we analyzed that metal to find out which components of the metal were critical to sound quality. We ended up coming up with a combination of four-nines tin [0.9999% purity] and another compound that’s a trade secret. It’s like cooking!

What other parts of a receiver get that kind of care?
One is the capacitors in the power supply. We specify the paper, the foil, the terminal, everything. For the new line of receivers, we listened to about 100 samples. Paper change, listen. Foil change, listen. Tension change, listen.

Did you specify the construction of this room?
Yes. When we came in, the original floor [the raised plywood floor commonly used in audio listening labs] they put in was too weak. I found a stiffer plywood that worked better, but we couldn’t get it because it was right after the earthquake [the 2011 Tohoku earthquake and tsunami that devastated the northern coast of Honshu, Japan’s largest island] and supplies of building materials were short. I ordered some from an American company, but the sound wasn’t as good because the Japanese factories use more pressure when they make plywood. So I ordered the wooden ply from America and had it laminated in a Japanese factory, and also added three more plies than the American factory was using.

What kind of system do you have at home?
I have two. My main system in my living room is a Sony STR-DA9100ES receiver; Harbeth Type 4 two-way speakers for front left, center, and right; ELAC 203 in-wall speakers for surround; small NHT models for front height speakers; and a Sony SA-WM500 subwoofer. That’s an inexpensive subwoofer, but for family use, it’s enough. I also have a Sony D50 three-gun projector.

You mean one of the old three-gun CRT projectors? Why are you still using that?
It’s not broken!

X