Interview with Peter Comeau, director of IAG

[Keith_W]

Peter Comeau recently visited Australia. Here he is giving an interview on Stereonet:

Some highlights that ASR might find interesting (my transcript, I have converted spoken English to written English, hopefully without changing the meaning. Any errors are mine):

04:35: "Every brand has it's own house sound, the reason for that is customer expectation."

08:45: "I am outspoken about measurements. I think there is far too much emphasis placed on measurements on the customer and reviewer side. I am sorry to say, there are some reviewers and a lot of comment on social media and forums about measurements. Measurements are a really good tool for the designer, they are not a really good tool for the analyst. Yes, I use measurements, I mainly use measurements to find out what is going wrong. [...] But it's not a good tool when you're trying to make a speaker that plays music, because measurements are based on steady tones, music is anything other than steady tones".

10:41: "Reducing distortion is something that we need to do. Not to incredible levels, I have been quite outspoken about this, in terms of pure harmonic distortion, it's very difficult to hear below 1%. So there is no point in chasing anything lower than that. And that explains why you can take a loudspeaker with harmonic distortion in the region of 0.5 - 1% and it sounds fine. This idea that people have with amplifiers, that you should be chasing 0.0001%, it's nonsense. What you should be paying attention to non-mechanical distortion in loudspeakers. In amplifiers it's caused by all sorts of electronic misbehaviour. And these things are things we can analyse using measurement and correct, but to do it on a reviewer basis, I think that's nonsense". <It's also very difficult to tell the difference between high bitrate MP3 and FLAC. Is he happy to settle for MP3 then?>

27:12: "In hifi we are not there yet [...] streaming, Class D amplifiers, the way we are learning to handle digital media, it's still in its infancy. The reason people like the sound of turntables is because we haven't quite got there in terms of the way we handle digital media yet." <No explanation given as to how digital media is "not there yet">

31:00: "Why have we got stuck in this era of frequency response and harmonic distortion, signal to noise, i.e. dynamic range. None of those things tells us anything about how music comes out of equipment. Why are we concentrating on just those specifications, instead of looking at the real problems with hi-fi?". <So what are the real problems in hi-fi?>

There is another interview with Darko where he says some controversial things about cables, more stuff about measurements, and so on. I'll post it when I find it.

It would have been nice if the interviewer had challenged him on some of these claims. The interviewer looked as if he was only reading off a transcript of questions to ask (to be fair to him, that was probably what he was asked to do).

 

[Phorize]

1)Measurements are a really good tool for the designer, they are not a really good tool for the analyst.

This pre-supposes that the designer and analyst (which I take to mean reviewer) aren't measuring for the same reason- which is not a good assumption.

2) measurements are based on steady tones, music is anything other than steady tones.

There is no reason that I can see to assume a 'steady' tone is less complex to reproduce than randomly variable multitones once one is working within certain mechanical and electrical parameters. He seems to be positing the existence of some hither too un theorised type of distortion or noise. It must be one of the two unless there are other audible signal properties that i don't know about.

He's right that we are well into diminishing returns on amplifier and distortion and noise. The innovation is as much about miniturisation and cost though. It must be hard these days to differentiate in the amplifier market, so I'd expect a ceo of IAG to introduce some FUD on this subject. However anyone who thinks that consumer speaker manufacturers consistently achieve basic best practices on speaker design isn't paying attention.

Of course, we can't ever rule out the possibility ofscientific advances in the future challengimg our conception of the world, but that's no more of a reason to dismiss the current state of audio engineering than it is the state of any other field. If we take his view, why not just casually dismiss our understanding of everything else in the universe.

 

[DSJR]

I used to know Peter Comeau as a supplier when he was at Heybrook and he HAS designed and 'voiced' some really nice speakers over the decades. I'd agree having spoken to a good few speaker designers over the years that LISTENING is vital once the basic tech design is done, as some things (I'm sure measurable) can be more annoying to our hearing than others and a DB tweaking here and there can be most effective without changing the basic measured performance.

Mind you, does the Klippel system use steady tones and sweeps? I gathered it uses 'chirps' of various frequencies, but I could be totally wrong here...

To answer the 'digital vs analogue' thing, one would have thought he'd have been in studios at various times actually listening to live vs 'recorded' playback. Even in the 70's the playback monitors easily reproduced the losses in getting said live performance onto tape and playing it back in direct comparison (drums especially). So many engineers have subsequently stated that the digital playback (before any signal manipulation in the workstation) is identical to the live feed...

When I had really decent (for the time) active speakers, the sound from a good digital source was an excellent if smaller scale facsimile of the real event and I never felt the need to compare, unlike in previous and subsequent times where the gap is much larger...

 

[Mart68]

seems to me his skill is in designing competent loudspeakers on a tight budget, I wouldn't question that he knows what he is doing there.

But I think he started out as a salesman and in his current position it isn't a good commercial move to say 'Oh yes, ASR is right.' Much more effective to use the opportunity to spread a little FUD, as someone else said here recently, educated customers are not something the hi-fi industry wants or needs.

 

[AC1]

Interesting! Peter Comeau says that measurements are only useful when designing a speaker but nonsense when reviewing a speaker.

 

[Kachda]

Interesting! Peter Comeau says that measurements are only useful when designing a speaker but nonsense when reviewing a speaker.

 

[TheBatsEar]

Right.

I liked the guy, but now he's on probation.