Modern Measurement Tools Are Tricking Audiophiles Into Trusting Bad Data, Warns Veteran Speaker Designer

[Sokel]

I watched him talk about designing that driver. Not a word about computer simulation in advance. Indeed, he says he had to build a prototype, not knowing at all if it would work or not. But got lucky that it did. A person schooled in the mathematics and physics of the topic we are discussing, would be simulating the design and optimizing it in that domain before anything physical is built.

Agree in general, it saves time and recourses obviously.
But as we traditionally know coaxials are usually a disaster unless the designer has a deep knowledge.

I mean one can be lucky, but not Mofi SourcePoint 8 kind of lucky, and I mean overall as a driver (that and a cabinet is all about it) AND cheap on top of it.
We have to be fair. Not my cup of tea as I like bigger stuff but what's fair is fair.




Edit: link.
Edit2: spin

 

[Blueprint]

I assume he heavily leant on his work developing the KEF UNI-Q driver.

 

[amirm]

Agree in general, it saves time and recourses obviously.
But as we traditionally know coaxials are usually a disaster unless the designer has a deep knowledge.

I mean one can be lucky, but not Mofi SourcePoint 8 kind of lucky, and I mean overall as a driver (that and a cabinet is all about it) AND cheap on top of it.
We have to be fair. Not my cup of tea as I like bigger stuff but what's fair is fair.

Edit: link.

No disagreement on that. We are not disputing his skills in that regard. The topic of this thread is whether something is faulty in the way Klippel NFS generates its results. If JA has criticized that, then we are saying his expertise in that area is limited and for sure, can't be used as an authority. I am pretty sure I know more about how Klippel works than him. But I can't remotely design that driver.

 

[Sokel]

No disagreement on that. We are not disputing his skills in that regard. The topic of this thread is whether something is faulty in the way Klippel NFS generates its results. If JA has criticized that, then we are saying his expertise in that area is limited and for sure, can't be used as an authority. I am pretty sure I know more about how Klippel works than him. But I can't remotely design that driver.

He does not criticize the machine or the results, he criticizes the way some people, sometimes, interpret the results.
For good reasons I reckon.

If as you say he, with physics degrees and all this experience (I would only wish to sit on his desk at TAD for a day! ) does not understand Klippel imagine the rest of us who not only can design a decent speaker pretty much so the stratospheric task designing a nice coaxial as his Kef or TAD or Mofi, etc.

 

[amirm]

He does not criticize the machine or the results, he criticizes the way some people, sometimes, interpret the results.

Why would that be different for any other measurement? Or is he claiming that we put aside all measurements for that reason?

 

[voodooless]

He does not criticize the machine or the results, he criticizes the way some people, sometimes, interpret the results.

So basically he found Twitter/X… welcome to reality!

 

[amirm]

If as you say he, with physics degrees and all this experience (I would only wish to sit on his desk at TAD for a day! ) does not understand Klippel imagine the rest of us who not only can design a decent speaker pretty much so the stratospheric task designing a nice coaxial as his Kef or TAD or Mofi, etc.

You are underselling yourself. Many of you know more about measurements than vast majority of audio designers out there! I routinely watch videos or read writing from key designers in audio companies, say wrong things that many of you would know just as well. You are getting exposed to measurements and how they work day in and day out. Many of those designers are not or they would already be investing in these instruments.

 

[Sokel]

Why would that be different for any other measurement? Or is he claiming that we put aside all measurements for that reason?

From what I got from the presentation, his argue is about small variations who can make a difference as a +2dB at 4kHz for example who can separate a nice speaker from a chainsaw (that's my vice I admit) and how we don't pay attention to such details.

 

[RickS]

Seems this already troubled thread cannot get back on topic. Giving it a rest and will consider whether it merits reopening.

 

[RickS]

Most of the off topic discussion here was due to continued trolling by one member. After reviewing the off topic content, I thought it had some great member responses to his posts. So rather than simply delete them, they have been moved to the “Are Measurements Everything or Nothing” thread. Also, rather than having one member’s misbehavior close the thread, will allow further discussion to resume here. Please note previous thread guidance and try to stay on topic. Thanks!

 

[René - Acculution.com]

No disagreement on that. We are not disputing his skills in that regard. The topic of this thread is whether something is faulty in the way Klippel NFS generates its results. If JA has criticized that, then we are saying his expertise in that area is limited and for sure, can't be used as an authority. I am pretty sure I know more about how Klippel works than him. But I can't remotely design that driver.

Indeed. There are some skilled people here that have a good grasp of various topics (what is a minimum phase transfer function, sound propagation, ...) at a higher level than many industry (not just loudspeaker) engineers.

 

[Gary_G]

While 95% of this thread was over my head I soo enjoyed the technical discussions, trying to pick up little tidbits here and there.

 

[ssashton]

I’m not sure if it’s worth reigniting this topic, but I can’t help but want to point out that the article never actually said that Klippel measurements have a 2–3 dB error. It only stated that AJ performed some measurements himself, then sent the device for independent testing and observed a variation of that amount.

That’s not unexpected. We don’t know the exact test conditions in either case, but differences of that scale can easily arise from factors such as the anechoic chamber, the measurement microphones (especially at frequency extremes), mounting conditions, drive equipment, and so on.

It’s true that many people misunderstand how the Klippel NFS works and misinterpret all kinds of data by taking it at face value without considering the testing context. However, the subtle implication that the Klippel NFS itself is inaccurate or is to blame for those misinterpretations, seems to me, almost certainly to be article bait.

 

[voodooless]

So, that whole argument is based on one anecdote that is at least 20 years old... And then he just says, and I paraphrase: well, nothing changed since then.

In reality, plenty changed, even if we do not include the NFS. The KEF anechoic chamber, as it was in 1994 when he left, was vastly different from the one they use now:

- It's much bigger
- There is a far higher angular resolution
- Full robotized automation

All of this makes the results way more reliable and repeatable, and as a consequence, also much more comparable to other sources that have a similar kit.

 

[rdenney]

Do any of these guys own anything other than t-shirts? If you want to be taken seriously, dress the part.

T-shirts and long (gray) hair are part of the look. One can't be a guru while dressed like the parents of the ex-protest-era hippies that are buying the high-end stuff these days.

Rick "ex-protest-era...well, that need not be said" Denney

 

[krabapple]

Most of the off topic discussion here was due to continued trolling by one member. After reviewing the off topic content, I thought it had some great member responses to his posts. So rather than simply delete them, they have been moved to the “Are Measurements Everything or Nothing” thread.

aka 'the cornfield'.

(I kid. Thanks for reopening this)

 

[John Atkinson]

The topic of this thread is whether something is faulty in the way Klippel NFS generates its results. If JA has criticized that, then we are saying his expertise in that area is limited and for sure, can't be used as an authority.

AJ (Andrew Jones), not JA.

John Atkinson
Technical Editor, Stereophile

 

[rdenney]

Just a note to add that someone may be extremely educated and qualified in an area, and still purposely avoid any discussion of their technical approach in a media interview, if that hasn't been the key to their branding. Amir's branding is all about measurement, but Mr. Jones's branding is more guru-oriented, and maybe he doesn't want to expose his engineering methodology in a media interview. I have no idea one way or the other, but I would not make assumptions about his methods based on what he says publicly about them.

His Pioneer BS22LR speakers, at a hundred bucks for the pair, are pretty decent. He had to make some compromises, of course, but they were reasonable and for the near-field application I use them for they are pretty good. I think there was some serious engineering in those speakers--I doubt he would have wanted to spend two years doing trial and error for the money Pioneer was paying him.

Rick "doesn't always disclose his own methods" Denney

 

[kyuu]

Just a note to add that someone may be extremely educated and qualified in an area, and still purposely avoid any discussion of their technical approach in a media interview, if that hasn't been the key to their branding. Amir's branding is all about measurement, but Mr. Jones's branding is more guru-oriented, and maybe he doesn't want to expose his engineering methodology in a media interview. I have no idea one way or the other, but I would not make assumptions about his methods based on what he says publicly about them.

His Pioneer BS22LR speakers, at a hundred bucks for the pair, are pretty decent. He had to make some compromises, of course, but they were reasonable and for the near-field application I use them for they are pretty good. I think there was some serious engineering in those speakers--I doubt he would have wanted to spend two years doing trial and error for the money Pioneer was paying him.

Rick "doesn't always disclose his own methods" Denney

True, but AJ has also had some duds. Like basically all his coaxial designs before the Spurcepoint as far as I can recall.

 

[Sokel]

True, but AJ has also had some duds. Like basically all his coaxial designs before the Spurcepoint as far as I can recall.

If TAD was his as well they are very good as well and if I got it right they are similar to Sourcepoint.