KEF R3 OR BOWER & WILKINS S706 s2 ???

[test1223]

No. No data. Anecdotes, stories and like they say opinions are as a certain body part. Everyone has one.

What did you think the word Science in the forum’s title meant?

Dismissing results of listening tests (even very flawed ones) as anecdotes is as unscientific as dismissing measurements with the argument to be only numbers.

Both measurements and listening tests can provide meaningful information as well as being insignificant, misleading or being not repeatable.

The state of the art predicted listener preference score is fare away from being perfect. About 2 points difference in the predicted preference score is needed to form a meaningful confidence interval. Which shows a good correlation but even without outliers there is a significant margin of difference between predicted preference score and preference score during a listening test.

The only test which can be seen as revealing "the truth" is a controlled listening test with all parameters in place. This test is very time consuming and the results might be only true for the exact side conditions.

 

[Mnyb]

Hmm remeber that lot of people also preferred loudness button on all the time and smiley face eq way back and speakers placed willy nilly ?

I like the idea of accurate speakers not pre equed to some target in a attempt second guess my my music taste ?
Which seems like what B&W are up to ? It might very well work splendidly for some tracks but not all ?

With this reasoning one needs a different speaker for each track you have .

An accurate speaker can sound bland and uninspiring to begin with but wins in the long term I think

Silly enough loudness and eq or tone controls are missing on high end equipment? So bake it in to the speakers instead ?

We can ofcourse have different taste about playback and actually not prefere accurate sound on all our music.

What makes this tread a little heated is that some B&W speakers have an intentional house sound that many prefer , but at the same time they fans can’t acdept that it’s not an accurate representation of the programme material.

It’s the weird notion of purity in the signal path that permitted high end for to long ?

So as purist you can’t have the needed tools to adjust tonality to your liking and shortcomings in the recordings and also have to pretend that your goal ofcourse is accurate sound aka HiFi ? And then you find a speaker that have your preferences for some of your music built in ? And mfg sells the idea that this is somehow accurate sound ?

Me I relaxed and turn up the bass when needed

 

[Galliardist]

OK, so let’s take a few other listening test anecdotes into account. For example, how do we explain the body of listening tests that claim that using multiple audiophile Ethernet switches improves streaming sound? By the same argument these tests must have value and the products must work in that way.

???

 

[HarmonicTHD]

Dismissing results of listening tests (even very flawed ones) as anecdotes is as unscientific as dismissing measurements with the argument to be only numbers.

Both measurements and listening tests can provide meaningful information as well as being insignificant, misleading or being not repeatable.

The state of the art predicted listener preference score is fare away from being perfect. About 2 points difference in the predicted preference score is needed to form a meaningful confidence interval. Which shows a good correlation but even without outliers there is a significant margin of difference between predicted preference score and preference score during a listening test.

The only test which can be seen as revealing "the truth" is a controlled listening test with all parameters in place. This test is very time consuming and the results might be only true for the exact side conditions.

Scientific data are quantified and replicable. Individual preferences (listening experience) by nature do not meet these requirements.

Scientific based listening tests (as you describe them correctly as controlled and statistically relevant) can therefore be generalized within the scope of their validity. Individual preferences (experiences) not.

The OP gives the impression that his individual preference (experience) has some general validity (that is how I perceive his post at least) and that is the problem.

Personally I avoid posting about my listening experiences at least not without clearly saying that those experiences are only valid for me in a specific room and were not controlled.

 

[witwald]

...multiple consistent observations about a loudspeaker in different environments by different individuals, albeit uncontrolled

In the above scenario, how likely is it that "consistent observations" can be achieved, and regarding which parameters?

In any case, subjective "observations" of loudspeakers often come from "influencers", usually accompanied by imprecise and emotive language describing the so-called "results". There is usually little or no data provided about any of the listening environments, the individuals involved, or the lack/presence of controls, let alone any objective measured data that may help explain things. Clearly, then, it's more important to just write the words than it is to do any measurements.

In a YouTube lecture by Al Bartlett, there was a statement that seems to apply to the situation here: "The more optimistic the prediction, then the greater is the probability that it is based on faulty arithmetic or on no arithmetic at all." If we apply Bartlett's observation to the situation at hand, then it would seem highly likely that subjective "observations" could be routinely faulty, as they quite commonly lack the underpinnings of any form of mathematical analysis.

 

[witwald]

The only test which can be seen as revealing "the truth" is a controlled listening test with all parameters in place. This test is very time consuming and the results might be only true for the exact side conditions.

So, if the results of controlled listening tests "might only be true for the exact side conditions" (which should of course be documented), what does that say about the usefulness, or otherwise, of "observations about a loudspeaker in different environments by different individuals, albeit uncontrolled"?

 

[test1223]

Scientific data are quantified and replicable. Individual preferences (listening experience) by nature do not meet these requirements.

No. Even data which contains only a very tiny part of "the truth" and is otherwise noise or biased can be used e.g. in social science and quantum mechanics it is done all the time.

The processing and interpretation is the most important part!

You simply claim that all listening test which aren't well conducted or documented are so flawed that there can't be derived a meaningful result with it. But your claim isn't backed by research. Especially if there are bigger differences, even flawed listening tests are more likely to be good enough and therefore "can" provide the same conclusions as accurately conducted listening tests.

So, if the results of controlled listening tests "might only be true for the exact side conditions" (which should of course be documented), what does that say about the usefulness, or otherwise, of "observations about a loudspeaker in different environments by different individuals, albeit uncontrolled"?

That these tests might contains useful information but might also be misleading you simply have a very high level of uncertainty. But the same is true for the predicted preference score. The level of uncertainty might be smaller but is it definitely not very unlikely that there is a different outcome if you conduct an actual listen test.

You also have to be aware that there is no second source which reproduced the research results of that predicted preference score (at least none that I am aware of). I think the research is valid at least for some specific side conditions but in other fields of research there is much higher skepticism until the results where reproduced by another group. There is also a lack of research under which side conditions the predicted preference score is more or less accurate.

The research here isn't completed but it isn't an unknown territory either.

 

[Galliardist]

No. Even data which contains only a very tiny part of "the truth" and is otherwise noise or biased can be used e.g. in social science and quantum mechanics it is done all the time.

The processing and interpretation is the most important part!

You simply claim that all listening test which aren't well conducted or documented are so flawed that there can't be derived a meaningful result with it. But your claim isn't backed by research. Especially if there are bigger differences, even flawed listening tests are more likely to be good enough and therefore "can" provide the same conclusions as accurately conducted listening tests.


That these tests might contains useful information but might also be misleading you simply have a very high level of uncertainty. But the same is true for the predicted preference score. The level of uncertainty might be smaller but is it definitely not very unlikely that there is a different outcome if you conduct an actual listen test.

You also have to be aware that there is no second source which reproduced the research results of that predicted preference score (at least none that I am aware of). I think the research is valid at least for some specific side conditions but in other fields of research there is much higher skepticism until the results where reproduced by another group. There is also a lack of research under which side conditions the predicted preference score is more or less accurate.

The research here isn't completed but it isn't an unknown territory either.

The problem starts before interpretation though. You need to define a problem, or present a hypothesis for experimental testing, that has a method to solution that can use that data.

Care to suggest something? The only thing we may learn is more about how people choose and evaluate components, as far as I can see. Even then, we’d probably need to interview and ask for a lot more data than what gets written in forum posts and so on.

 

[Ageve]

OK, so let’s take a few other listening test anecdotes into account. For example, how do we explain the body of listening tests that claim that using multiple audiophile Ethernet switches improves streaming sound? By the same argument these tests must have value and the products must work in that way.

???

Or the "quantum stickers":

Quantum Sticker - Telos Audio Design

Jeff Lin is the operator and owner of Telos Quantum technique.

telos-audio.com.tw

telos-quantum-x2-stickers - GrooveWorks Aust Pty Ltd

TELOS Quantum X2 Stickers (sheet of 20)

recordclean.com.au

"Peter, I can’t believe what I’m hearing! This is incredible!"
- G.C. Tasmania (Quantum Stickers)

"This is insane! I can’t think of any logical reason why these stickers should do what they do…but they’re doing it anyway…and how! This is the best, most cost-effective upgrade I have ever tried...and that's over 40 years in audio! If this is audio voodoo, give me more of it!"
- P.S. Victoria (Quantum X2 Fuses and Quantum Stickers)

"This sceptic is well and truly convinced…I’ve gone dotty!"
- B.L. Victoria (Quantum X2 Fuses and Quantum Stickers)

"If I had found these things years ago I would have saved a lot of money. I’m hearing things I’ve never heard before!
The ambiance, the imaging, the voice, the definition…amazing!
The noise floor dropped dramatically! I can hear everything! The best upgrade in audio!"
- P.W. NSW (Quantum X2 Fuses and Quantum Stickers)

"I have never heard anything like this…anywhere!"
- D.L. NSW (Quantum X2 Fuses and Quantum Stickers)


If we are going to turn this into Audio Subjective Review, my brother owns a pair of KEF R5s, and I have listened to the 706 S2 as well. I find it very strange that anyone would prefer the B&Ws. To me, they sound bright and tinny, and the soundstage is lacking (I currently own a pair of Revel F208 and M16).

Subjective opinions based on uncontrolled listening are, well... not reliable.

Just look at this review of another B&W speaker, the 606 S2 (5/5):

Bowers & Wilkins 606 S2 Anniversary Edition review

What Hi-Fi? Awards 2022 winner. B&W hones its 606 standmounters to great effect

www.whathifi.com

Move up the frequency range and the steps up in articulation and clarity are striking. Voices come through with greater subtlety, and it’s easier to hear changes in intonation and phrasing. The Series 2 offers a clearer view of the recording and sounds more balanced overall.
...
Their presentation is tauter, and a little more forward than the originals, but also sounds more natural and transparent. Tonally, there’s been a shift away from what now seems like a slightly over-blown bass and overly rich lower midrange to something more neutral.
...
As the music builds, these speakers have the composure and organisation to keep things under control. They rarely sound flustered or stressed no matter how demanding things get.
...
Along with the improved muscle, we’re finding it easier to track low-level instrumental strands and pick up on fine detail. The new model conveys instrumental textures with more skill and locates instruments within the sound stage with greater stability.


And then you have this review (It's the same speaker. S2 is the Anniversary Edition. And yes, I know he liked the 705 S3).

The review is very negative, and so are the comments:

I just got my B and W 606 s2s up and running today. I was shocked. It's for this reason that I went online to see if others have had the same experience as mine. Well, I agree completely with the reviewers here. These are very harsh speakers. There is 'bright' and then there are the 606 S2s. These are being packed up and returned tomorrow.
...
Totally agree, i have those speakers and it is so bad that I thought they where broken
...
I have the 606 and I can't enjoy music on these speakers. make me fatigue/headachy. the same songs even sound more musical on my Huawei p30 phone speakers. it is definitely harsh in my experience and I hope to replace it in the future
...
I got some B&W 706s and they sound sort of OK but a bit ordinary for the price.The voice sound isn't near throaty enough and sounds just a little bit muffled while the highest highs are stinging. The ride cymbal rocks but the voice sounds 2D and "boxed in".
...
I have owned a pair for 2 months and have tested them with what I consider two be warmish amplifiers and I although they do some things well they are to bright. They give me a headache after an hour. I've actually bought a cheap pair of Q acoustic 3020 as a place holder until my 606s2 sell on eBay.
...
And I thought my hearing was damaged because most people were praising these speakers
...
I love B&awesome but hated these. After 15 minutes I was totally fatigued. So glad to read a review that echos EXACTLY how I found them.
...
I bought the 606 s2 because I enjoyed the previous 606s so much,but I have to agree with you on the 606 s2.they are all treble and hardly any bass,and I sold them within 2 weeks of buying them
...
Guys, I couldn’t agree more with you on this review. I actually bought a pair of 606 s2 anniversary editions and found the sound beautiful but fatiguing. It was honestly bittersweet to return them because they are special speakers minus the crazy brightness issue, but even after hours of burn in I just didn’t want to listen to music.
...
I'm in my 60's and I found these speakers too damned bright and not neutral when I heard them. And I hate excess sibilance! I shun such speakers (and am just as repeled by one-note bass as well)! Thanks for the confirmation; wow!
...
Everyone who disagree with this don't have good and sensitive ears or are fooling themself. Doesn't matter what amp you are using or what kind of room. They sound just terrible anyway.


But there are B&W fans as well:

Must be something wrong with your ears?
...
Obviously can't get these speakers free from B&W hence this nasty video
...
This review officially puts you in the fool category. Can't watch anything you put out again.
...
Its for the views. They wanted to get 200k fast and to do that, shitting on some well known product is the obvious thing to do. They lost my respect and unsubscribed.


So what did we learn from all this, without looking at measurements? Nothing, really. It's all just subjective "data".


From the Reddit thread:

Sales guy in the room stated a strong preference for the BW 706 S2 vs the R3.
…
the BW sounded significantly better. To my untrained ears, the 706 S2s just seemed to have more clarity, more 3d imaging, just more overall ”enveloping” sound. The music coming out of the speakers just seemed noticeably ”richer” to my ears.
…
R3 where on a shelf, only a few inches from back wall.
…
706 s2 were on a shelf as well, but due to the size difference of the speakers, they had more room behind them.
…
Source material was just what was on Best buys Demo application screen — a few live DMB songs and some other stuff I wasn’t as familiar with.


To summarize, a random guy/girl with "untrained ears" walks into a Best Buy. The sales guy states a strong preference for a bright speaker, and they listen to stuff that the customer isn't familiar with. The speakers are on a shelf - not stands, and the KEFs are placed "only a few inches from the back wall".

I mean, who needs measurements, when we have "data" like this?

B&W speakers are designed to "impress" at a demo. The result is pretty much expected.

 

[test1223]

I was not aware that we were discussing quantum mechanics or the social sciences here, but it might be interesting if you could take the time to expand on your statement.

Please explain how we can derive a meaningful result from any test which is not well-conducted or documented. After all, on the face of it, this seems to follow closely the Subjectivist apologia, which assigns a great deal of value to points of view not supported by logic.

Jim

The problem starts before interpretation though. You need to define a problem, or present a hypothesis for experimental testing, that has a method to solution that can use that data.

Care to suggest something? The only thing we may learn is more about how people choose and evaluate components, as far as I can see. Even then, we’d probably need to interview and ask for a lot more data than what gets written in forum posts and so on.

My recommendation is: don't dismiss the possibility that the B&W speaker is actually better. A high amount of sales and according reports have to have an explanation. The research of the predicted preference score doesn't falsifies the hypothesis that the B&W speaker is better!

I mentioned social science because everyone can at least to some extent relate to that and there you even have to deal with claims which can't be verified or falsified.

How do you verify news articles or rumours? You usually can't do that, therefore you assume how likely different scenarios are which could have generated the article or rumour. You deal with hypothesis, clues and possibilities.

 

[Steve Dallas]

I have. And who's to say that a treated room in a hifi shop is going to match YOUR living room better than the BB listening room? In fact, in reality, many people need to place their speakers in non-ideal locations to blend in with furniture/aesthetics/etc.

It is a common group think concept that you can't trust any listening evaluations that occur in Best Buy. That is nonsense. Free your minds of group think. Imagine if that post was about someone who compared the R3 and the B&W in a BB showroom and confirmed that the R3 was superior. Would folks here object and say that the B&W's probably sounded better and the BB environment caused a faulty comparison? Let's be honest here.

And... You are projecting yet again.

There is no group think. I have been to several BBs with Magnolia showrooms for auditions. Here is what I found each time:

1. Electronics switcher injects noise and distortion into the signal chain, which is enough to mask the performance of any electronics in the chain. This also makes it impossible to evaluate speakers for distortion. Full on buzzing ground loops at one location
2. Haphazard speaker placement
3. Very high environmental noise floor
4. Employees barely know how to operate the system and know next to nothing about equipment and speakers, other than higher price + brand = always better

The above make visits to BB / Magnolia pointless for me. The environment is NOTHING like my home environments.

The BB argument is just a distraction, however. What matters is, the assertion is a single anecdotal data point that can be added to our collective memories of the subjective data set. [shrug]

I don't know what people would say if the reverse were the case. I can only assume people would say the KEF was chosen even in the cacophony that is BB / Magnolia. My own thought is that it is not terribly surprising that showroom sound won out in a terrible showroom. [shrug]

Anyway, I have stated my thoughts on this comparison and provided measurements to support my reasoning and subjective opinions. I'll not waste any more time on this. If people like that sound, more power to them. I will take accuracy on every day ending in 'y' instead.

 

[preload]

And... You are projecting yet again.

Are you sure you understand what the term "projecting" refers to? Pretty sure you don't based on how you seem to be applying the term.

There is no group think. I have been to several BBs with Magnolia showrooms for auditions. Here is what I found each time:

1. Electronics switcher injects noise and distortion into the signal chain, which is enough to mask the performance of any electronics in the chain. This also makes it impossible to evaluate speakers for distortion. Full on buzzing ground loops at one location
2. Haphazard speaker placement
3. Very high environmental noise floor
4. Employees barely know how to operate the system and know next to nothing about equipment and speakers, other than higher price + brand = always better

The above make visits to BB / Magnolia pointless for me. The environment is NOTHING like my home environments.

If they're using an active switcher that is causing audible distortion, then you would clearly be justified asking them to remove it from the chain. If you're nice to people and not there to disparage the employees and their showroom, I can't imagine why they wouldn't let you.
And we're talking about an untreated room with a high environmental noise floor, that's EXACTLY like the home environment - vs a sound treated demo room in a hifi store.

The BB argument is just a distraction, however. What matters is, the assertion is a single anecdotal data point that can be added to our collective memories of the subjective data set. [shrug]

Please re-read. Collective data points are useful.

I don't know what people would say if the reverse were the case.

They would be supportive (if the anecdotal report were that KEF > B&W). That's my point. That's the group think. In research, that conclusion would be called "confirmation bias."

Anyway, I have stated my thoughts on this comparison and provided measurements to support my reasoning and subjective opinions. I'll not waste any more time on this. If people like that sound, more power to them. I will take accuracy on every day ending in 'y' instead.

If you like sound that you think is "accurate," then great, more power to you. I like sound that sound reproduction that sounds "preferable" to me, and that is the basic premise of the Harman research too.

 

[preload]

In the above scenario, how likely is it that "consistent observations" can be achieved, and regarding which parameters?

Consistent: as in multiple, independent individuals reporting that they "preferred" the sound of the B&W to ____ speaker.
Again, the occurrence is, at the very least, hypothesis generating.

In any case, subjective "observations" of loudspeakers often come from "influencers", usually accompanied by imprecise and emotive language describing the so-called "results".

I don't watch those videos (don't have the patience), and I'm also not referring to them. I'm simply referring to the fact that in most threads here that mention B&W, there WILL be people who mention that they preferred XYZ B&W model to another loudspeaker. And these are folks responding in a B&W-hostile environment. And on other forums on the interweb, there are a ton of people who commonly prefer the B&W sound. And of course, we know that several renowned recording studios also use B&W in their monitoring rooms.

I also personally own (or have owned) B&W's, Genelecs, Kefs, and Revels, and I quite frankly, the group think idea that B&W's cannot sound good because of the measurements is patently false.

 

[preload]

I mentioned social science because everyone can at least to some extent relate to that and there you even have to deal with claims which can't be verified or falsified.

While admittedly not the point being tried to made here, I'll also add that the Harman research correlating loudspeaker preferences to loudspeaker measurement analysis is MUCH CLOSER to social science research than it is to the discipline of engineering.

 

[preload]

Or the "quantum stickers":

Quantum Sticker - Telos Audio Design

Jeff Lin is the operator and owner of Telos Quantum technique.

telos-audio.com.tw

telos-quantum-x2-stickers - GrooveWorks Aust Pty Ltd

TELOS Quantum X2 Stickers (sheet of 20)

recordclean.com.au

"Peter, I can’t believe what I’m hearing! This is incredible!"
- G.C. Tasmania (Quantum Stickers)

"This is insane! I can’t think of any logical reason why these stickers should do what they do…but they’re doing it anyway…and how! This is the best, most cost-effective upgrade I have ever tried...and that's over 40 years in audio! If this is audio voodoo, give me more of it!"
- P.S. Victoria (Quantum X2 Fuses and Quantum Stickers)

"This sceptic is well and truly convinced…I’ve gone dotty!"
- B.L. Victoria (Quantum X2 Fuses and Quantum Stickers)

"If I had found these things years ago I would have saved a lot of money. I’m hearing things I’ve never heard before!
The ambiance, the imaging, the voice, the definition…amazing!
The noise floor dropped dramatically! I can hear everything! The best upgrade in audio!"
- P.W. NSW (Quantum X2 Fuses and Quantum Stickers)

"I have never heard anything like this…anywhere!"
- D.L. NSW (Quantum X2 Fuses and Quantum Stickers)


If we are going to turn this into Audio Subjective Review, my brother owns a pair of KEF R5s, and I have listened to the 706 S2 as well. I find it very strange that anyone would prefer the B&Ws. To me, they sound bright and tinny, and the soundstage is lacking (I currently own a pair of Revel F208 and M16).

Subjective opinions based on uncontrolled listening are, well... not reliable.

Just look at this review of another B&W speaker, the 606 S2 (5/5):

Bowers & Wilkins 606 S2 Anniversary Edition review

What Hi-Fi? Awards 2022 winner. B&W hones its 606 standmounters to great effect

www.whathifi.com

Move up the frequency range and the steps up in articulation and clarity are striking. Voices come through with greater subtlety, and it’s easier to hear changes in intonation and phrasing. The Series 2 offers a clearer view of the recording and sounds more balanced overall.
...
Their presentation is tauter, and a little more forward than the originals, but also sounds more natural and transparent. Tonally, there’s been a shift away from what now seems like a slightly over-blown bass and overly rich lower midrange to something more neutral.
...
As the music builds, these speakers have the composure and organisation to keep things under control. They rarely sound flustered or stressed no matter how demanding things get.
...
Along with the improved muscle, we’re finding it easier to track low-level instrumental strands and pick up on fine detail. The new model conveys instrumental textures with more skill and locates instruments within the sound stage with greater stability.


And then you have this review (It's the same speaker. S2 is the Anniversary Edition. And yes, I know he liked the 705 S3).

The review is very negative, and so are the comments:

I just got my B and W 606 s2s up and running today. I was shocked. It's for this reason that I went online to see if others have had the same experience as mine. Well, I agree completely with the reviewers here. These are very harsh speakers. There is 'bright' and then there are the 606 S2s. These are being packed up and returned tomorrow.
...
Totally agree, i have those speakers and it is so bad that I thought they where broken
...
I have the 606 and I can't enjoy music on these speakers. make me fatigue/headachy. the same songs even sound more musical on my Huawei p30 phone speakers. it is definitely harsh in my experience and I hope to replace it in the future
...
I got some B&W 706s and they sound sort of OK but a bit ordinary for the price.The voice sound isn't near throaty enough and sounds just a little bit muffled while the highest highs are stinging. The ride cymbal rocks but the voice sounds 2D and "boxed in".
...
I have owned a pair for 2 months and have tested them with what I consider two be warmish amplifiers and I although they do some things well they are to bright. They give me a headache after an hour. I've actually bought a cheap pair of Q acoustic 3020 as a place holder until my 606s2 sell on eBay.
...
And I thought my hearing was damaged because most people were praising these speakers
...
I love B&awesome but hated these. After 15 minutes I was totally fatigued. So glad to read a review that echos EXACTLY how I found them.
...
I bought the 606 s2 because I enjoyed the previous 606s so much,but I have to agree with you on the 606 s2.they are all treble and hardly any bass,and I sold them within 2 weeks of buying them
...
Guys, I couldn’t agree more with you on this review. I actually bought a pair of 606 s2 anniversary editions and found the sound beautiful but fatiguing. It was honestly bittersweet to return them because they are special speakers minus the crazy brightness issue, but even after hours of burn in I just didn’t want to listen to music.
...
I'm in my 60's and I found these speakers too damned bright and not neutral when I heard them. And I hate excess sibilance! I shun such speakers (and am just as repeled by one-note bass as well)! Thanks for the confirmation; wow!
...
Everyone who disagree with this don't have good and sensitive ears or are fooling themself. Doesn't matter what amp you are using or what kind of room. They sound just terrible anyway.


But there are B&W fans as well:

Must be something wrong with your ears?
...
Obviously can't get these speakers free from B&W hence this nasty video
...
This review officially puts you in the fool category. Can't watch anything you put out again.
...
Its for the views. They wanted to get 200k fast and to do that, shitting on some well known product is the obvious thing to do. They lost my respect and unsubscribed.


So what did we learn from all this, without looking at measurements? Nothing, really. It's all just subjective "data".


From the Reddit thread:

Sales guy in the room stated a strong preference for the BW 706 S2 vs the R3.
…
the BW sounded significantly better. To my untrained ears, the 706 S2s just seemed to have more clarity, more 3d imaging, just more overall ”enveloping” sound. The music coming out of the speakers just seemed noticeably ”richer” to my ears.
…
R3 where on a shelf, only a few inches from back wall.
…
706 s2 were on a shelf as well, but due to the size difference of the speakers, they had more room behind them.
…
Source material was just what was on Best buys Demo application screen — a few live DMB songs and some other stuff I wasn’t as familiar with.


To summarize, a random guy/girl with "untrained ears" walks into a Best Buy. The sales guy states a strong preference for a bright speaker, and they listen to stuff that the customer isn't familiar with. The speakers are on a shelf - not stands, and the KEFs are placed "only a few inches from the back wall".

I mean, who needs measurements, when we have "data" like this?

B&W speakers are designed to "impress" at a demo. The result is pretty much expected.

TL;DR

"Brevity is the soul of wit." -- Hamlet

 

[preload]

I don't. I don't believe anyone else here does, either. "Better", however, is based on subjective preference. "More accurate" (or not) is based on tests and measurements applied with discipline, rigor and logic.

Accuracy is interesting.
Accuracy for simple electrical devices that have a 2-D signal in and signal out can be quantified with measurements. Easy. DACs, amplifiers, sources, etc. can be quantified by SNR, DR, FR, etc.
BUT loudspeakers are not simple electrical devices. They are electromechanical. The input is a 2-D signal, and the output is a complex signal that varies in polar space and in the time domain, all of which further varies as a function of the playback room.

So how can you define "accuracy" in a loudspeaker via measurements? And more importantly, do those measurements actually correlate the subjective impression of "accuracy" by humans?

This dilemma of measuring "accuracy" in loudspeakers was most certainly recognized by the Harman research group. I believe this is one of the reasons why they chose to use PREFERENCE as their outcome of interest to measure instead of "accuracy."

 

[Ageve]

TL;DR

"Brevity is the soul of wit." -- Hamlet

Noted.

 

[preload]

No. No data. Anecdotes, stories and like they say opinions are as a certain body part. Everyone has one.

What did you think the word Science in the forum’s title meant?

Please explain how we can derive a meaningful result from any test which is not well-conducted or documented. After all, on the face of it, this seems to follow closely the Subjectivist apologia, which assigns a great deal of value to points of view not supported by logic.

Jim

Just because an individual data point wasn't derived in a controlled well-documented setting, doesn't mean that AGGREGATED data points derived in such a fashion aren't useful. Quite the contrary actually. There's a whole scientific research methodology based on this, and it allows us to study a variety of phenomena that are not amenable to a "laboratory" test.

 

[preload]

Measurements are the ONLY method for defining accuracy, in loudspeakers, in hydraulics, in chemistry, in metallurgy or in any other physical discipline that depends on controls and descriptions derived from scientific processes.
That is the reason that the phrase, "... subjective impression of 'accuracy' ... " is a non-sequitur.

Okay, thank you for clarifying. As you point out, many people use the term "accuracy" to describe a perceived subjective characteristic of a speaker. For instance: "speaker A sounds more accurate than speaker B"

BUT, for the sake of argument, I'm willing to roll with your definition. So let's define accuracy, for our purposes here, as defined by measurements only.

How, then, can you use measurements to say that Speaker A is more accurate than Speaker B?
Would you define "accuracy" as the deviation between the output and the input? (if not, how would YOU define it in terms of measurements)?
If so, then measurement of the speaker output would consist of recording the output signal of at all angles of radiation in 4-pi space. Perhaps you can perform some analysis of that, and determine the FR deviation (from the input signal) at all angles in 4-pi space. But then what? Do you then analyze the degree of FR deviation? How do you weight that? In 1/20th octave resolution? 1/3 octave? And do you weight various frequency bands higher because they are more important to "accuracy?" What about the spectral tilt of the FR curve, do you weight that? And so on and so forth.

How can you take all of that, and then use measurements to say that Speaker A is more accurate than Speaker B? And THEN what? When you say that Speaker A is mathematically more accurate than Speaker B, does it THEN mean that Speaker A will SOUND "subjectively" more accurate than Speaker B? (i.e. will a group of humans listening in blinded/controlled conditions agree that Speaker A sounds "closer to a reference source/speaker/condition" than Speaker B?

Again, I agree that from a practical standpoint, accuracy based on measurements is a practical and meaningful way to characterize electrical equipment with a simple input/output signal (like DACS, amps, etc.). But the practicality and interpretability seems to fall apart with transducers, which are complex electromechanical devices that take a simple amplitude vs. time signal and transform it into a signal that radiates into 4-pi space and varies based on the radiation angle as well as in the time domain (due to reflections). And all of this is assuming an ideal or standardized room. When playing back in different rooms, the measurements will differ once again.

So please tell us, how do you interpret loudspeaker measurements in order to objectively determine whether Speaker A is more "accurate" than Speaker B?

 

[Galliardist]

Accuracy is interesting.
Accuracy for simple electrical devices that have a 2-D signal in and signal out can be quantified with measurements. Easy. DACs, amplifiers, sources, etc. can be quantified by SNR, DR, FR, etc.
BUT loudspeakers are not simple electrical devices. They are electromechanical. The input is a 2-D signal, and the output is a complex signal that varies in polar space and in the time domain, all of which further varies as a function of the playback room.

So how can you define "accuracy" in a loudspeaker via measurements? And more importantly, do those measurements actually correlate the subjective impression of "accuracy" by humans?

This dilemma of measuring "accuracy" in loudspeakers was most certainly recognized by the efinitioHarman research group. I believe this is one of the reasons why they chose to use PREFERENCE as their outcome of interest to measure instead of "accuracy."

What is interesting though is that the research into preference led back to a logical definition of accuracy, a flat anechoic response. What’s more, the preferred headphone curve was defined from in-room measurements of a loudspeaker which closely matched that definition of accuracy.

It’s also interesting that the results are more consistent for the midrange. And that the preference breaks down when comparisons are uncontrolled.

None of this makes those brief subjective descriptions useful information unless we ask the right question of it.