Evidence-based Speaker Designs

[Ilkless]

The OP has a view which likely fits the saying "a little bit of knowledge is a dangerous thing." He understands some speaker design principles but not others, taking the view that any data regarding listener preference should be discarded. In his view the beloved BBC style monitors are crap because of their intentional frequency response variations. The OP's dogmatic view is not the reality of what people want to listen to.

There are some speakers which come close to his ideal like the Kii 3 and Genelec 8351. However, even those get an occasional tepid review. He would explain those reviewers are idiots. The simpler explanation is mathematical perfection isn't for everyone.

At Audiogon it isn't anti-intellectualism. They are calling a stone a stone.

Please do not distort my words. How can the denial of empirical evidence in favour of personal anecdote be anything but anti-intellectual?

Vaccines occasionally get tepid reviews. Their detractors are similarly calling a stone a stone. Do we condone it?

 

[Ron Texas]

Please do not distort my words. How can the denial of empirical evidence in favour of personal anecdote be anything but anti-intellectual?

I am not distorting anything. It's been pointed out that you are ignoring an important category of empirical evidence. You need to be asking questions about why does the LS50 sell while the similar speaker from Techics does not. The answer isn't people are stupid or KEF has a better marketing strategy.

 

[Ilkless]

I am not distorting anything. It's been pointed out that you are ignoring an important category of empirical evidence. You need to be asking questions about why does the LS50 sell while the similar speaker from Techics does not. The answer isn't people are stupid or KEF has a better marketing strategy.

Anecdote without the most basic controls (particularly to avoid bias of a non-acoustic nature) hardly constitutes evidence. I thought this was part of the ethos of this forum and has been covered extensively in numerous threads, so it is surprising to see that you would argue otherwise.

 

[Ron Texas]

Anecdote without the most basic controls (particularly to avoid bias of a non-acoustic nature) hardly constitutes evidence. I thought this was part of the ethos of this forum and has been covered extensively in numerous threads, so it is surprising to see that you would argue otherwise.

Now you are distorting what I and a few other participants in this thread are saying.

 

[Ilkless]

Now you are distorting what I and a few other participants in this thread are saying.

You said:

...any data regarding listener preference should be discarded. In his view the beloved BBC style monitors are crap because of their intentional frequency response variations. The OP's dogmatic view is not the reality of what people want to listen to.

There are some speakers which come close to his ideal like the Kii 3 and Genelec 8351. However, even those get an occasional tepid review. He would explain those reviewers are idiots. The simpler explanation is mathematical perfection isn't for everyone.

This implies you regard subjective sighted reviews as data of equal validity to what has been cited in this thread. Unless you mean to say there were in fact blinded, volume-matched "tepid" reviews of these speakers, which is extremely unlikely.

You also said that there was data about "listener preference" that provided an alternative basis of empirical evidence. In an earlier reply, I have already substantiated how there is a psychoacoustic basis to the Toole/Olive preference results. The preference ratings, read alone, are irrelevant, useless and non-generalisable until we relate them to how human hearing mechanisms generally operate in-room. I am open to peer-reviewed research that contradicts the evidence already offered, but it seems there has not been any in the replies.

 

[Ron Texas]

You said:

You are attempting to parse every word of mine in a useless semantic battle. Your problem is a complete inability to see the big picture. The unacceptable flaw you find in the LS50 is far more likely the result of science than sloppy engineering. The so-called BBC dip you obviously abhor is there because of an intentional design choice and it works as intended. The uneven frequency response seen in B&W speakers is intentional. Some say it gives the sound the illusion that it is layered.

No doubt there is a lot of in house research which isn't published because they are trade secrets.

 

[MZKM]

The simpler explanation is mathematical perfection isn't for everyone.

These comments weren’t done via double-blind, level-matched, quick-switching listening, so their reviews of such products are tainted by unconscious bias, the brand, price, and asthetic all play a role in our evaluations of sound quality.

Now, Toole’s research (and others) do show that not everyone likes the exact same amount of bass or treble, but that still doesn’t prove that measurements can’t tell us what speaker we will like better under the same parameters when listening under the conditions listed above. Of course, if comparing two speakers that are [near] identical except one has anechoically flat bass and the other is slightly boosted, then yes, measurements won’t tell us which one every human will like more, but most speakers are not that close in measured performances

 

[Ron Texas]

These comments weren’t done via double-blind, level-matched, quick-switching listening, so their reviews of such products are tainted by unconscious bias, the brand, price, and asthetic all play a role in our evaluations of sound quality.

However tainted, how can you be sure the result was not correct?

 

[Krunok]

I am not distorting anything. It's been pointed out that you are ignoring an important category of empirical evidence. You need to be asking questions about why does the LS50 sell while the similar speaker from Techics does not. The answer isn't people are stupid or KEF has a better marketing strategy.

Argument that something is a good product because many people are buying it is actually a false argument. It is called "argumentum ad populum" and is well explained on Wiki. I think you should read it.

 

[Ron Texas]

Argument that something is a good product because many people are buying it is actually a false argument. It is called "argumentum ad populum" and is well explained on Wiki. I think you should read it.

I don't see in that quote any assertion that the LS50 is good because people are buying it. My point is the OP is overly focused on his interpretation of what evidence is. He needs to look at why the LS50 sells and the Technics doesn't. Where there is smoke there is fire. Perhaps the answer is the LS50 is better or it might be something else. What is a good product? Are we stuck here that the definition of a good speaker is ruler flat frequency response? That used to be the approach at Consumer Reports. They ignored sensitivity, dynamic range, distortion and whatever else goes into the batch.

Seems you are suffering from the problem of a little bit of knowledge is a dangerous thing by thinking that Wiki article is the be all and end all. It completely missed the entire category of non fallacious reasoning in this area.

https://philosophy.lander.edu/logic/popular.html


"There are many non-fallacious appeals in style, fashion, and politics--since in these areas the appeal is not irrelevant."

" Conventional truth such as the definitions of words, standard use of symbols, and clothing styles, or voting in juries, meetings, or political elections are typical examples where the appeal to the majority , the experts, or the people-in-the-know would be relevant and so would not be fallacious.

"If an elite group of people are in a position to know of what they speak, their authority is relevant and should not automatically be discounted. E.g., Is is a legitimate appeal and no fallacy to argue that most physicians believe that a high fat diet is unhealthy, and therefore a high fat diet is unhealthy."

Gee wiz, I wonder if all those great reviews of the LS50 came from the elite? You probably think it's all BS because the tests were sighted.

 

[MZKM]

However tainted, how can you be sure the result was not correct?

Because one person like say it’s neutral and someelse says it’s harsh, and another saying it lacks PRAT, and anothing saying it lacks emotion.

In Toole’s double-blind studies, he states the better measured speaker is always chosen as the best. However, he is testing different speakers, and not different bass/treble amounts, which is the caveat I gave. His protege/co-worker Sean Olive did test frequency response preferences:

 

[JJB70]

I may be wrong, but I suspect that if you asked KEF's design engineers, or indeed the design engineers of most of the major speaker manufacturers, they'd be able to provide entirely rational and "evidence based" arguments in support of their design choices. Ditto the major AV producers and companies like Dolby despite the fact that hi-fi buffs seldom seem to want to acknowledge the contribution of home theatre to sound. I return to my point about Bose, any list of speaker manufacturers with a pedigree in research and analytical design must include Bose whatever we might think of the products they make.

 

[Ron Texas]

In Toole’s double-blind studies, he states the better measured speaker is always chosen as the best. However, he is testing different speakers, and not different bass/treble amounts, which is the caveat I gave. His protege/co-worker Sean Olive did test frequency response preferences:

Do you think the people who are buying Kii 3's have a median age of 32? It appears to say that some deviation from an absolutely flat frequency response is desirable and varied by group. I wonder how a bunch of decrepit fossils like me would test. Note that the average 55 year old male can't hear frequencies above 12 khz and it gets worse as we get older.

I have been practicing being able to exclude most frequencies of the female voice in an effort to preserve my sanity.

 

[MZKM]

I see a graphic detailing who was in the listening group, but no results. Do you think the people who are buying Kii 3's have a median age of 32?

It’s a short presentation off an AES paper, if you find/buy the paper, there likely is more data.

The older subjects liked less bass and less treble (until they reached 55+ where they didn’t mind more treble as they couldn’t hear that high anymore). Via the Stereophile’s measurements, the bass is not dampended and the upper treble is slightly recessed.

 

[Ron Texas]

I may be wrong, but I suspect that if you asked KEF's design engineers, or indeed the design engineers of most of the major speaker manufacturers, they'd be able to provide entirely rational and "evidence based" arguments in support of their design choices. Ditto the major AV producers and companies like Dolby despite the fact that hi-fi buffs seldom seem to want to acknowledge the contribution of home theatre to sound. I return to my point about Bose, any list of speaker manufacturers with a pedigree in research and analytical design must include Bose whatever we might think of the products they make.

It's unlikely that you are wrong and any other result would not be rational.

 

[Ron Texas]

It’s a short presentation off an AES paper, if you find/buy the paper, there likely is more data.

The older subjects liked less bass and less treble (until they reached 55+ where they didn’t mind more treble as they couldn’t hear that high anymore). Via the Stereophile’s measurements, the bass is not dampended and the upper treble is slightly recessed.

I wonder why? The over 55 response to treble is obvious, but why a general preference for a lack of bass. Must be us old men buying up all the stand mounts with their restricted low frequencies.

 

[MZKM]

I wonder why? The over 55 response to treble is obvious, but why a general preference for a lack of bass. Must be us old men buying up all the stand mounts with their restricted low frequencies.

Keep in mind that these preferred levels are higher than what a neutral speaker produces in-room (the room being Harman’s own with room treatment), and since those with more listening experience also liked less bass than those with little experience, a correlation most likely exists.

 

[Krunok]

He needs to look at why the LS50 sells and the Technics doesn't. Where there is smoke there is fire.

I'm really impressed with your "logic". McDonald hamburgers sell as well, so what a great food they must be.

I'm sure it all looks simple to you when you observing the world around you through your eyes and applying your "logic".

 

[RayDunzl]

McDonald hamburgers sell as well, so what a great food they must be.

I was so disappointed when McDonald's quit posting an estimate of the number of burgers sold.


"McDonald's Sign Milestones

1955- 1 million (Ray Kroc's first McDonald's opens in Illinois)
1956- 5 million
1960- 400 million
1963- 1 Billion (served by Ray Kroc himself on national tv)
1969- 5 Billion
1976- 20 Billion
1984- 50 Billion
1987- 65 Billion
1990-80 Billion

April 15, 1994- 100 Billion (McDonald's executives announce at the annual owner operator convention that they will stop counting hamburgers served because the count has surpassed the 99 Billion Hamburger mark so all the operators should change their signs to say "BILLIONS AND BILLIONS SERVED" and the signs have remained that way since."

 

[svart-hvitt]

Are there any speakers that aren't designed based on evidence? If you use alleged golden ears to design a piece of audio equipment it is still evidence based, just a different sort of evidence (one that I'd be very sceptical about).

To your question:

http://www.hifi-musik.se/extra-material/2018/12/Magnat-Transpuls-1500/

Take a look at the FR charts.

What design criteria went into those?