Bruno Putzeys Navigates Toward the State of the Art

April 2016

In my last column, I wrote about the wonderful Channel Islands E•200S two-channel amplifier. But what really got me excited was Bruno Putzeys, the Belgian designer of the Hypex UcD module used in the E•200S. Putzeys has been in the audio world for several years. He started at Philips, where he worked extensively in labs, experimenting with input stages, power types, and supplies. Unfortunately, as soon as he came up with a great design, he was confronted with a multinational conglomerate’s tendency to do nothing when presented with a new idea.

Bruno Putzeys

As you’ll read below, Putzeys is not solely motivated by a chase for the almighty euro. He is an unusually thoughtful person who wants to democratize quality, and is an unqualified supporter of using science as a major part of engineering. When Philips threw too many roadblocks in his way, Putzeys jumped ship for another Dutch company, Hypex.

Founded in 1996 by Jan-Peter van Amerongen, Hypex originally made plate amplifiers to stick on the back of studio monitors or PA systems, to convert them to active. The amplifiers themselves became quite popular, and Hypex came to a turning point: Should they make parts for other companies to use, or should they make their own components? They chose the former and, with Putzeys as their new chief of R&D, began selling amplifier modules.

In the considerable time I spent talking with Putzeys on the phone and e-mailing back and forth, I got the impression that I was dealing with an honorable man who is loyal to his friends and business partners, and a defender of the general public from all sorts of advertising bunkum. Not that Putzeys described himself in this way -- but, as I learned in grad school, always trust actions over words.

Putzeys always tried to ensure that I didn’t slip in anything that might embarrass his associates. For instance, in asking which major companies use Hypex modules, I guessed that one of them was a popular maker of ultra-high-quality professional monitors. He told me that even though it’s an open secret, he wanted to honor their wish that he not mention it. He could have slipped it in, or asked me to mention it but not let them know that he’d told me. Instead, he made a straight request that I leave it out.

Putzeys speaks a number of languages -- his American English is virtually accent free. He’s genial, generous with his time, quite smart, and often very funny in an ironically mocking vein. He is appropriately modest and when he steps outside the bounds, it’s less braggadocio than a challenge to have someone prove him wrong. For example, in the realm of power amplifiers, it’s clear he believes that he’s made as good an amp as can be made, which is why he’s moving on, to speakers. Rather than bloviate about his amps, it felt more like been there, done that. Putzeys believes that he and his partner, Bart van der Laan, can improve the state of the art of loudspeaker design using what he’s learned in his years in electronics. The professional world is paying attention. When the Kii Three speaker was launched, it sold out almost immediately. But faced with far more orders than they could easily fill, Kii changed exactly nothing -- no cutting back to make production faster, no reduction of circuit quality. Instead, they apologized for “badly underestimating how many orders we would get.”

. . . nicht weniger als ein Geniestreich. Warum hat das zuvor noch keiner gemacht?

So says Wolfram Eifert in Germany’s Audio magazine. Just in case your German is as rusty as mine (this I memorized: Kann ich bitte ein Hefeweissbier?), here is what this all means -- and you know you must be around someone truly special when you keep hearing the same words everywhere: “nothing less than a stroke of genius. Why hasn’t anyone done this before?” This refers to the Kii Three speaker, but it could have been about Putzeys’s Hypex amplifiers, his Grimm Audio speakers, his Mola-Mola amplifiers, or the multitude of professional monitors that use Hypex parts for their active circuitry.

Putzeys’s sky-high reputation is something of an enigma, because he takes every opportunity to launch a firebomb at the equipment-review trade. He says that they yammer -- that “one of the joys of audio writing is generating baroque prose to describe the sonic vagaries.” On the Hypex site, the new and improved UcD module is described thusly: “Distortion (THD and IMD) over the full audio and power range is negligible, typically below 0.0007%. Distortion at listening levels (1W) is unmeasurable. This amp adds neither dirt nor fairy dust.” I can’t help but think that Bruno Putzeys may have had something to do with that description.

This interview took place late January through early March 2016. When Bruno Putzeys thinks the emperor has lost his clothes, he seems incapable of not pointing it out.

Wes Marshall: It’s clear to me is that, somehow, you understand something that the powers that be have either never understood, have avoided for some reason, or were afraid of.

Bruno Putzeys: Don’t just do philosophy. Make a mental list of the issues that you want to solve, and then go looking for a solution. Most audio companies have a core philosophy, like using a particular cone material, or use only FETs, or no feedback, etc. -- something that defines their identity. That hampers any serious attempt at solving problems. If, for mere reason of ego or identity, you’re not free to use any means available to solve a problem, you’re not going to do a good job. A speaker drive unit is not some magical machine to make sound, with an inherent sonic character, etc. It’s a device for moving air. It’ll have some secondary unwanted effects, so pick the one whose unwanted effects you can work around.

But I take it you’re also referring to creativity. I guess I’m quite creative. Creativity, whether in the arts or in engineering, consists of finding a solution that is particularly effective, preferably at solving several problems at once or, better still, avoiding them. Schools typically teach you to pick a problem apart into smaller problems, solve each in turn using precooked recipes, and then put these partial solutions together. Well, that guarantees a monster every time.

All I can say is: Try to study as many relevant things as possible; i.e., collect a big but vague cloud of knowledge around the problem. Be greedy for knowledge. On the other hand, make sure you keep your design brief sharply in mind. Go out walking and biking often, hang out with friends. But don’t bother consciously trying to attack the problem. What you then find is that your subconscious occasionally pings you to ask if you could verify a particular reasoning step, and when you do, it goes under again. Then, suddenly, it comes back up, saying, “Well, you could get from there to there, from there to there, etc.” -- spelling out a solution that works but is not necessarily elegant. So now you have to work on it consciously again to try and find shortcuts. But the mere fact that you now know a solution exists is what you need to find the elegant solution.

WM: Given the fact that class-D was almost universally derided, it took something of a renegade, iconoclastic, fully committed person to choose to make it. Why D?

BP: Partly for moon-shot reasons: not because it’s easy, but because it’s hard. I’ve always had a sense that I wanted to make a real difference. For me, a career in class-AB amplifiers would be as interesting as being a bank teller. If you want different, don’t start at same. Why class-D, specifically? There is exactly one good reason to want to do class-D: to save energy. Apart from this outstanding benefit, a class-D amp will initially try to do just about anything, bar amplify music. The technical challenge of taming it is huge, and pretty much defines the fun of the enterprise.

Energy saving is good for the planet, although I’m under no illusion what percentage of global energy use this will affect. The size, however, is about to turn the complete audio scene on its head in the form of active speakers. The separates audio industry is moribund, and class-D is cheerfully digging the grave.

WM: Which companies can we say use your Hypex amp modules?

BP: I have to name names with care, because some of those folks are technically competitors of my own brands, Mola-Mola and Kii. So I try not to keep too much abreast of who has designed-in the modules that Hypex sells. Off the top of my head, the first names that pop up are NAD, Jeff Rowland, MBL, Channel Islands, and Auro 3D’s surround processors. The list is much longer, because in the past few hi-fi shows I’ve done, there were always three to four other companies using my designs in just the same corridor. Quite a few pro speaker companies, too, but I guess that only Barefoot will be somewhat known in consumer circles. Kii too, of course.

Mola-Mola KalugaMola-Mola Kaluga mono amplifier

WM: So far, other than Mola-Mola, who is doing the best job with the Ncore modules you designed for Hypex?

BP: In terms of value, I think the Acoustic Imagery Atsah 500 monoblock is a really good implementation of the Ncore. [Note: $1995 for two amps, 400W into 8 ohms, 700W into 4 ohms.]

WM: When I was checking for some biographical information about you, I came across this statement, posted on in February 2012: “With to him good reason, Bruno does not believe in anything audiophile. If you can measure something it’s there and if it’s there you can measure it. That’s his credo.” Is that true?

BP: It’s not a very precise way of stating my position, but it’s close. What I’m saying is that we have to pull audio away from fairyland and back into engineering. The audiophile world became suspicious of measurements when, in the ’70s and ’80s, it became common practice to sell amps based only on THD at 1kHz and 1W, and listeners found that that number had nearly zero correspondence with perceived sound. That’s obvious. It is entirely possible for an amp to have a really good THD at 1kHz and 1W and have absolutely useless performance at 10kHz, 10W. In fact, if that single number is all you’re measuring, you’re rather likely to end up with a design like that. Audiophiles definitely had a point: using a single data point to specify an amp doesn’t work, and is terribly bad engineering. But then to conclude we can just as well do away with measurements entirely is plain bad logic. This notion that measurements don’t say all has bloomed into a catchall excuse to sell the most preposterous items and ideas. It’s exactly the same as paranormal peddlers saying, “There are more things between heaven and earth,” etc. -- which sounds like a call for an open mind, but they’re really using it to say that if you can’t give a cogent explanation for something that they find odd, their pet faerie theory is thereby confirmed. Can’t measure the effect of microscopic dots ($300 a pop) on the wall? But you can hear it, no? Sure, P.T. Barnum, I hear you, and I take note of your clever use of suggestion.

I’m saying: We need more measurements, not fewer. If you can hear a change in the sound that comes out of your speakers, it means that the electrical signal that’s gone in has changed. And if you clearly hear a change that you can’t measure, how about trying different measurements? Sweep the frequency band. Sweep the level range. Try multiple signals at once. Break down the distortion into its constituent parts. By then, you’ll see tons of changes. Try to correlate those with your listening experience.

And learn to think critically about listening. If the device you’re working on has line-level ins and outs, set its gain to 1 and build a bypass switch to hear whether it changes the sound or not, instead of comparing two boxes and picking the one you prefer. Because if both change the sound, then, clearly, comparing measurements is only going to confuse you.

And please, learn how you can be fooled. Tell a friend you have found this amazing trinket, fiddle around behind his stereo without actually changing anything, and then watch him explain precisely what differences he’s just heard. Make sure you have an excellent bottle of Scotch on standby to make up for it.

And to my fellow engineers, I’d almost like to say the reverse: Take people’s listening experiences seriously. If they hear things (outside a prank or a sales pitch) that you can’t explain, don’t explain it away, but go looking. It’s amazing how often I’ve ended up working out a new measurement trick because a few people have independently given the same feedback.

To make a long story short: I am convinced that, for everything that is provably audible, a measurement can be devised -- and, furthermore, it’s quite clear to me that most supposed mysteries are already solved.

WM: Already solved? That’s a statement that should send fear into a lot of amplifier manufacturers. Given the entrenched interests in the world of manufacturing, not to mention R&D, I’m quite curious. Anyone who makes a discovery with enough power to shift the normal paradigm is bound to get pushback. At bare minimum, they’ll be branded as crazy -- perhaps even a death sentence will be issued.

BP: I think that the Ostracized Genius is largely a creature of mythology. People love stories of patents of water-powered cars withering away in some oil giant’s strong room (apparently forgetting that a patent is a publication), or of inventors of cold-fusion devices who find that every company they approach has just been visited by men in black pointedly suggesting that certain inventors aren’t to be talked to. These myths are mostly perpetuated by cranks who like to paint themselves as a contemporary Galileo. It isn’t until you start bumping against actual political ideologies that any such risk develops. In reality, if an oil giant realizes you’ve got a pocket fusion device that really works, they’ll invest in it. Better to be an ex-oil fusion giant than an ex-energy giant.

That doesn’t mean that the ones who don’t get it won’t fight a rear-guard battle. Somewhere in the early 2000s I wrote a long Usenet post about so-called “digital” amplifiers -- that is, amplifiers that try to drive their power FETs directly with a digital signal instead of building a proper analog error controller. Immediately, I got a request from a Canadian manufacturer of high-end class-AB amplifiers if they could put the post on their site. Which they did, believing that the article somehow dissed class-D in general. No, it didn’t. In fact, it clearly hinted at a class-D amp that actually behaved like an amplifier. It was a teaser to attract attention to UcD, and it was hosted for years on the site of a class-AB amp company hoping it would dissuade people from going class-D.

WM: Something that keeps coming up in our discussions is your strong aversion to people who make outrageous claims for products -- strong enough that you feel some responsibility to point them in the right direction. Did any company you worked for or school you learned from enhance this sense of fair play? Or is this purely a “Bruno” thing?

BP: Steven Pinker (Johnstone Professor of Psychology, Harvard University) makes a strong case that people’s character is in large part genetic, with the remainder not correlating much with upbringing. I’ll accept his findings. All I know is that, on the whole, I’m quite left-leaning. Also, over time, I found I was fairly easily taken advantage of by People With Plans. I’m trying to be much more careful about that now, but equally, it riles me enormously when someone else is getting suckered.

WM: How would you describe the sound of your amplifiers?

BP: It’s interesting that when one compares my amps to run-of-the-mill solid-state amps, the first impressions that pop up are “more valve-like mid/highs.” A remarkable number of tubeophiles have said that, with Ncore, they can finally let go of valve amps. I think that means that the beef they had with the sound of the average solid-state amp was valid. Indeed, if you compare Ncore against a seriously good solid-state amp, you’ll find that, in fact, it sounds very, very similar. In other words, these same qualities are shared with other highly evolved designs. It’s always a good sign when different amp technologies converge to the same sound. When designers forgo spurious “audio philosophies” and just get on with the job of building an amp with as few as possible demonstrable flaws, they are bound to get similar results. That “sound,” or lack of it, is the neutrality and transparency I seek.

WM: Do these “audio philosophies” always hurt the sound? Isn’t it possible that they might be on to something?

BP: I would say there’s almost a moral duty to try and keep personal tastes out of audio design. It is a matter of respect to artists not to “remaster” their work. Euphonic modifications are fair game on the production end when artists are in charge, but the reproduction end should not add further commentary to that. Commercially, too, it makes sense. Unless, for some strange reason, everyone has the same biases as oneself, a neutral sound will, on the whole, gain you most friends. And, of course, as I already intimated, neutrality meshes well with the engineering approach. If you design for absolutely flawless measured performance, the result turns out to be sonically neutral too.

WM: I play a little bad guitar. Do you have any reasoning in mind as to why professional guitar players almost all use tube amplifiers? I know that when I play one, especially if I play loud and staccato, it seems to respond more quickly to the plectrum and my fingers. In other words, it seems to pop -- if we were recording from the speaker, it seems to have a steeper, taller waveform -- in a way that most non-tube amps can’t replicate. If that’s true, is that what attracts consumers to tubes for home stereo? With Hypex modules, I seem to find that the sound just exists on its proper plane, with no sense of compression or expansion.

BP: The reason why valves sound good on guitar is because those amps aren’t designed for low distortion. The amp and the cabinet are designed together to form a part of a musical instrument. They pretty much replace the chamber and soundboard. They’re designed by a musical-instrument maker, so it’s no wonder they’re perfectly suited for making music. If you use a hi-fi system or studio monitor to listen directly to the output of the guitar, you have only half of an instrument. So I don’t think the usefulness of valves in musical instruments has too much import on their popularity in hi-fi, although I’m quite happy that, thanks to the hi-fi industry, guitarists can now get decent modern valves, besides a dwindling supply of NOS valves.

In the case of home stereo, I really think that there is something many solid-state amplifiers do wrong that valve fanatics aren’t willing to put up with. The fact that they do tend to like Ncore suggests that everyone just has certain sonic idiosyncrasies they’re willing to tolerate in order not to suffer some of the others, and that those are different for everyone. It’s more about what they don’t like rather than what they do. But present people from opposite ends of the spectrum with something that just sounds right [and] they’ll almost always agree. I’d say the majority of audio lovers are looking for the same thing, but they’re approaching it from different angles.

WM: Do you play any instruments? And what do you listen to when you have the odd free moment?

BP: I’m afraid that every time I considered taking up an instrument, I geeked myself out of learning to play. So, sadly, no -- I don’t play. Worse still, like most audio engineers, I end up listening to music way too rarely, for lack of time. Musically, my ground zero is what’s loosely called “world music”; more precisely, various indigenous traditions as existed before pop/rock came and stomped all over it. Which is not to say I don’t like fusion -- in fact, I love it -- but the traditional component really should be more than a thin veneer over an R&B core. Probably the best-known example of music that truly managed to pull together various musical influences and make it into something unique is the Ethiopian jazz that was played, late ’60s, early ’70s. Sadly, there are virtually no reasonably good recordings [of it] around. Otherwise, I’m fairly omnivorous. Classical, rock, R&B, electronica -- anything, so long as it isn’t cheap.

WM: Do you feel that a Mola-Mola system driving a set of Kii Threes would be the current state of the art for Bruno Putzeys? Is there anything coming that would advance that state? If yes, anything you might be able to either describe or at least hint at?

BP: Well, the Kii Three is a digital active speaker, so unless you need a turntable input, you wouldn’t use external electronics. In my home, where I don’t have much in the way of room treatment, the Kii Threes are the only speakers that really work -- which was, of course, the idea. That said, the Dutch Vivid [Audio] distributor has promised to stop by and optimize the setup of my Vivid [Giya] G3s, so I hope this will help level the playing field a bit. It’s probably no secret that, among passive speakers, I hold Vivid Audio in high regard, as well as YGA. Both companies seek neutrality and work hard at reducing nonlinear distortion in their drive units, another subject that I am keenly interested in. Lowering distortion in loudspeaker drivers, by whatever means, is clearly the next frontier.

WM: I agree with you on that. It’s amazing to me how few speaker makers want to discuss specifications of things like distortion. It seems as if some genius use of DSP would be key. Shifting your career to speakers from amp modules, an area we can perhaps say you have already conquered, is a huge change. How is that transition going?

BP: Segueing to speakers actually felt quite natural. In my time with Hypex, I helped speaker companies design our modules and DSP units into their products. To get a sense of what would be required from a turnkey DSP-amp unit, I started replacing the passive filters in my own speakers with whatever DSP prototype I was working on. I’d never thought about speakers much, but what surprised me was how many aspects of speaker design were plainly obvious from my signal-processing background that were entirely new to seasoned speaker designers. It really turned out that the signal-theory stuff I’d trained myself to do on amplifiers and DSP was a fast track to understanding speaker design. When I tried passing those insights to our first DSP customer, they politely rebuffed me, explaining that if I do the amplifiers, they’ll do the speakers, thank you very much. Well, all right -- that wasn’t particularly polite, so while I doubt they intended to throw down the gauntlet, that’s the effect it had. My first commercial speaker design was the Grimm LS1, which came out three years afterward. The second is the Kii Three, which has just started shipping this winter.

But yes -- there has, of course, been a “big moment” in all this, and that’s when I went to tell Jan-Peter [van Amerongen, who runs Hypex] that I wanted to wind down my work for Hypex in order to concentrate fully on loudspeakers. Europeans on the whole are not very entrepreneurial, so the whole changeover, leaving Hypex and starting Kii, is quite an education.

Kii ThreeKii Three loudspeakers

WM: I was looking at the Kii Three’s specs and was amazed at one of them. It states a frequency response of 30Hz-25kHz, +/-0.5dB. How is that possible? How did you measure it?

BP: Anechoic, on axis. I should clarify that spec. The speaker is within 0.5dB of a second-order Butterworth high pass at that corner frequency, so the low corner is -3dB -- but the deviation from flat over the remainder of the band is 0.5dB.

WM: Thanks!

Putzeys has several fans among the writers at SoundStage! If you’re interested in the basics of negative feedback, see “Searching for the Extreme: Bruno Putzeys of Mola-Mola, Hypex, and Grimm Audio -- Part One” and “Part Two.”

And if your engineering chops are pretty advanced, you can try Putzeys’s “The F-word, or Why There Is No Such Thing as Too Much Feedback,” and his “Ncore Technology White Paper.”

. . . Wes Marshall