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

[pma]

Sure. If I throw a stone in a pond, you see a bunch of circles expanding

A 3D pond and a pulsating sphere . Just to make the analogy even more real.

 

[Curvature]

@amirm There are two things I'm quite curious about.

When the Klippel software computes the estimated in room response, what's the height of the speaker? Specifically, I mean do the computed response take into account that a bookshelf rests on a stand while a floorstander stands on the floor?

Following that, what's the placement of the speaker and the room like in that computed response? Obviously, the "room" has no nodes, but what kind of parameters does it use for that imaginary room?

Estimated in room response in a spinorama presentation is an average of other curves. It only uses anechoic data.

 

[pma]

Here is the formula from NTK thread:

BTW, the origin is from Wolfgang Klippel’s publications. NTK has borrowed it.

 

[Arindal]

The best is probably the KH80, which was measured by @amirm three times using multiple units, and also measured by Erin, @napilopez, @Nuyes, Anselm Goertz at Sound & Recording. ...Despite all that they align well with Neumann's own comprehensive measured data, and Neumann themselves reviewed @amirm's work.

Not surprised at all. A very compact active 2-way monitor with a waveguide-loaded tweeter, steep x-over and frontal reflex vents, is the category of speaker I would always expect the lowest margin of error from between different measurement methods.

those from anechoic chambers, which are typically only rated as such to 100-200Hz, below which a compensation curve is applied to measurements.

Serious lab engineers would always choose a different method other than compensation, if their anechoic chamber does not allow free-field conditions for longer wavelengths. Adding true nearfield measurements of the sound sources in question, with an individual method of summing them up and calibrating to the free-field part, is the preferred way, some add groundplane measurements. This does not work with automated routines, though, always requires the work of an experienced specialist.

 

[Curvature]

Serious lab engineers would always choose a different method other than compensation, if their anechoic chamber does not allow free-field conditions for longer wavelengths.

Standards dictate certification, and where a chamber can't meet the standard it is not certified. The reporting states this clearly. Compensation, nearfield or groundplane (in hemianechoic rooms) are all normal, but compensation is the most common because it is quickest and reasonably accurate.

Not surprised at all. A very compact active 2-way monitor with a waveguide-loaded tweeter, steep x-over and frontal reflex vents, is the category of speaker I would always expect the lowest margin of error from between different measurement methods.

Like I wrote above, the bigger issue is the availability of measurements, not accuracy.

Even in the case of other more complex speaker designs, I've not seen evidence of inaccuracy. Lots of speculation, but no evidence.

 

[Heinrich]

Comment/opinion from Andrew Jones UK loudspeaker designer. I think he has some valid points to consider in loudspeaker measuring.

This site is not one I follow closely I rarely use headphones.

Robert

Tl;dR, sorry. But reading the thread diagonally, I didn't find a comment on how to rank the different measurements from the Klippel in regard to their importance.
- direct on-axis, that point in space is rarely hit by two (!) ears simultaneously
- listening window, my ears don't average a field, they are at one place at a time
- predicted, a standard room doesn't exsist

Toole says, on-axis is it for the most part, while with healthy designs the steady state (ref 'predicted') will follow nicely w/o further consideration. Healthy is no abrupt changes in directivity or even reversal.

The latter is a clear caveat against 'predicted', because for example a highly reflective frontwall plus less reflective backwall will sound difefferent than low refl/ front combined w/ high refl/ back. I've seen some fellows discussing eagerly half a dB (!!) in the directivity plots ...

And still, the listener is not even part of the equation. Where is he, on his cosy sofa forget about stereo triangle and head orientation, or on his critical listening chair, right height, eyes straight, keep it!?

We have the data, but how to use it other than predicting (other peoples') preference?

 

[Keith_W]

Tl;dR, sorry. But reading the thread diagonally, I didn't find a comment on how to rank the different measurements from the Klippel in regard to their importance.
- direct on-axis, that point in space is rarely hit by two (!) ears simultaneously
- listening window, my ears don't average a field, they are at one place at a time
- predicted, a standard room doesn't exsist

I think we have to read the Klippel measurement intelligently. The data it gives you does not apply in your room. But it does give a reasonable prediction of what will happen in your room. Nobody is saying that the PIR is what you will get at home. I always ignore the PIR (if there is a good reason to pay attention to it, I am open to changing my position!). Why bother with the PIR when you can take a measurement at home and give you the AIR (Actual In-Room response)?

As for "listening window", it will tell you how large your sweet spot is going to be. I don't sit still while I am listening, I move around, eat cheese, browse for more music, etc.

 

[Curvature]

Tl;dR, sorry. But reading the thread diagonally, I didn't find a comment on how to rank the different measurements from the Klippel in regard to their importance.
- direct on-axis, that point in space is rarely hit by two (!) ears simultaneously
- listening window, my ears don't average a field, they are at one place at a time
- predicted, a standard room doesn't exsist

Toole says, on-axis is it for the most part, while with healthy designs the steady state (ref 'predicted') will follow nicely w/o further consideration. Healthy is no abrupt changes in directivity or even reversal.

The latter is a clear caveat against 'predicted', because for example a highly reflective frontwall plus less reflective backwall will sound difefferent than low refl/ front combined w/ high refl/ back. I've seen some fellows discussing eagerly half a dB (!!) in the directivity plots ...

And still, the listener is not even part of the equation. Where is he, on his cosy sofa forget about stereo triangle and head orientation, or on his critical listening chair, right height, eyes straight, keep it!?

We have the data, but how to use it other than predicting (other peoples') preference?

Read Toole's book for detailed answers to all those questions.

I always ignore the PIR (if there is a good reason to pay attention to it, I am open to changing my position!

PIR and AIR match above 1kHz.

 

[DanielT]

I don't think so. My experience with AJ is, he is both an objectivist and subjectivist at the same time and genuinely trusts his listening test results as much as his measurements, believing that both camps should take advantage from the other one´s findings.

Could not agree more with him. Hearing his TAD R1 almost two decades ago, was the decisive moment for me to realize what really differentiates a great sounding speaker from one which just does avoid the common mistakes.

It is possible that you are right. Everyone who is or claims to be an objectivist experiences, subjectively, the same things differently due to preference, experience, mood... there is much more to add, but you get the point.
Having said that, I would not, for example, even consider any speakers that showed a terrible result on all conceivable objective measurement criteria. At least not any expensive, "end game" speakers.

I have bought a lot of cheap used speakers on the spur of the moment to test them, because it is fun. Some gems, a lot of crap, which were then sold fairly quickly.

OT:
Do you think my hypothesis about the very friendly, cuddly and fluffy language: Atkinsonian is correct?
(Not as bad as the Swedish magazine Hifi Musik, hell what ass-kissing to their advertisers, (which always got highly praised reviews), they indulged in.)

 

[Multicore]

Well, nothing he's saying is exactly wrong. But what or who exactly is he railing against? If he has an issue with a particular reviewer, he should come out and say it (and provide his reasoning).

Well, exactly. This stuff doesn't wash in real critical discourse. You have to cite your references.

It's also a well-know straw man of the "People think/say/do X, which makes them wrong and me right" type. A variant of this common in journalism softens the insult a little with the "We tend to believe X but actually Y" type, where the author patronizingly includes him/herself. I call that the "Malcolm Gladwell We".

 

[Heinrich]

Read Toole's book for detailed answers to all those questions.

Read "the book"--well, I did. The answers are not in the book. Hence the questions reiterated.

The listener, what do they do actually, when not used in a controlled environment as probes for speaker quality? Dunno if Andrew Jones really said what was insinuated, but to step back once in a while and scrutinize the latest achievements was real 'science'.

Only half serious, take away the A versus B element of Harman's testing. Just one speaker, no comparison. That's where most consumers land in the end. There is no better or evenen different anymore. Question is, is it good enough? Is this questioned answerd, seriously? If I was more a music enthusiast, not an audiophile ...

 

[Curvature]

Read "the book"--well, I did. The answers are not in the book. Hence the questions rereiterated.

The research question was: is it possible to use only anechoic measurements of the speaker to predict in-room response and listener ratings despite all of the complexities involved? The answer was: yes.

 

[Somafunk]

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

On the contrary, the biggest shysters and shillers of snake oil come wearing suits n' shirts.

What someone wears, or chooses not to wear in these modern times has very little bearing on expertise.

The idea that to be taken seriously you need to dress the part disappeared up its own arse decades ago

 

[NTK]

I’m fascinated by the NFS’s ability to calculate far field response from close mic measurements in the context of the DUT being a large multi-way speaker where layman understanding suggests that correct/intended driver summation doesn’t occur until the measurement position is far enough away, something like 2-3x largest driver spacing. Is there a qualitative explanation you can help with as I highly doubt I could understand the maths (which I found hard enough when taking my acoustics degree 25 years ago!). Thanks.

The theoretical basis of Klippel's NFS is near-field acoustic holography (NAH), which is one of the most important advances in acoustics of the past 4 decades. Measuring near-field can therefore be an important advantage, not a liability.

Four decades of near-field acoustic holography

The Reflections series takes a look back on historical articles from The Journal of the Acoustical Society of America that have had a significant impact on the

pubs.aip.org

One of the inventors of the technique, Dr. Earl Williams, wrote a book (https://doi.org/10.1016/B978-0-12-753960-7.X5000-1) that has become a widely used text for university acoustics classes (for example: See MIT's graduate level Advanced Structural Dynamics and Acoustics class).

The basis for the sound field separation technique was from Professor Weinreich's 1980 paper.

https://pubs.aip.org/asa/jasa/article-abstract/68/2/404/781387/Method-for-measuring-acoustic-radiation-fields?redirectedFrom=fulltext

The specific technique that is used by the Klippel NFS in its computations is Professor Wu's Helmholtz equation least squares method (HELS). The method is explained in details in his book (https://doi.org/10.1007/978-1-4939-1640-5).

To summarize, the theoretical foundations of Klippel's NFS is totally solid.

 

[Sokel]

On the contrary, the biggest shysters and shillers of snake oil come wearing suits n' shirts.

What someone wears, or chooses not to wear in these modern times has very little bearing on expertise.

The idea that to be taken seriously you need to dress the part disappeared up its own arse decades ago

I'm wearing a pink one writing "I'm with the f*cking band" right now, do I need to take it off to comment?

Seriously now, whatever suits anyone, as long as there is no code (written or unwritten)

To the thread, AJ does not need credentials, its speakers talk for him (both measurable and subjectively) .
Apart from its Kef stuff (which I know very little, I was never interested in Kef) whatever else I have listened from him has the very first quality to me:

It's not strident and has meat. This 2-3dB he's talking about is what takes speakers apart to me (given the same size, etc of course) .
Take that from the upper bass and put it at the 1kHz-5kHz and you have the equivalent of a small chainsaw.

Given the matter of accuracy we're talking about here it's not easy to see such small but crucial stuff. That's why one should listen after establishing the basics.
Specially the ones who are sensitive in sibilants as they have turn to an epidemic these days.

 

[a4eaudio]

...I'm not sure I can agree or disagree, since I do not know this person--either through reputation or personally.

Chris

Google him. One of the premier speaker designers on the planet. But, that still doesn't mean he's right about everything.

 

[Heinrich]

The research question was: is it possible to use only anechoic measurements of the speaker to predict in-room response ...

This is trivial, do you see that? Otherwise the presence of room (or listener) would alter the speaker's output. O/k, with the Klipsch horn that might be the case, for sake of the argument. But in general, it's acoustic wave mechanics, isn't it?

... and listener ratings despite all of the complexities involved?

This is impossible. You cannot predict my personal preference--times I don't like any speaker because it makes some noise, that simple. Conversely, all the ratings depend on an unspecified mindset of the audience, if we nonchalantly go statistics to begin with. The mindset when sitting in "that room" at Harman's may differ from that of a couch potato. The latter type is more common than audiophiles tend to consider.

I'm with no word against objective data, I appreciate "the standard" as such fullheartedly. But it can and has been overstreched far beyond. By audiophiles of the other kind, true enthusiasts, but still a little bit misled, me thinks.

Remember, what if there is no A versus B comparison anymore? How much of a 'defect' remains objectionable, especiall on the long run? Settle and enjoy, that might be a new mantra after virtual perfection is available (by standard;-), courtesy of Toole/Harman. Andrew's designs seem to follow that, big bass, big volume, not textbook linear, but: Good Enough!

 

[Curvature]

It's fairly clear Andrew Jones has nothing to do with that article and his comments are taken out of context.

 

[vsch]

@amirm — I have a question that may be naive, as I lack formal training in signal processing (though I have some limited knowledge of econometrics).

When using the Klippel NFS, do you need to specify how many drivers a speaker has and the distances between them, or can the system infer that information purely from the measurements it takes?

Put differently, are the solutions the NFS computes always unique, even if you don’t provide prior information about the sources of the different wave components?

Thanks so much for the tremendous public good you provide through this site and all the detailed reviews — needless to say, it has been game-changing.

 

[RickS]

As several members have already questioned the credibility of the article referenced in the OP, will share some supporting info...

1. The article's author is the same person who wrote the (apparently now retracted) article about the fragility of KEF drivers. So not too surprising if the article has some significant subjective bias...
2. While the article seems to acknowledge the value of measurements, it places way too much value on our hearing. While see no issue with questioning sound quality based on one's hearing, it is less reliable as a sole reference than measurements. If it sounds really bad, likely there is good reason to investigate further. When it gets to minutiae, hearing is more suspect and should be suitably scrutinized. Otoh, measurements deserve scrutiny as well, but as many here know, more accurate usually is better and more of a reason to question what you are hearing and NOT the other way around.
3. I just banned one headphonesty spammer. Along with some recent questionable content from them, if you post content from them here, do not be surprised if it gets some added scrutiny. As with all offsite content, the requirement to provide a summary is partly to give some pause before posting it here. Notably, considering the source should be seriously done beforehand!