KEF Reference3 vs b&w 804 D3

[<eargiant]

BTW, I didn't get any blackness crazy details when I had my B&W 802D2 using the ATI AT3002 amp and the Denon AVP-A1HDCI pre-pro.

Did you get any blackness crazy details when you listened to the 803D?

But I guess everyone hears differently.

Maybe you had them set-up incorrectly??

Here's what one first-time listener had to say about the 802D3:

"I was just pretty blown away at how realistic they sounded."

"I wouldn't change a thing. They really do sound awesome."

"They were very articulate, realistic and there was plenty of clean powerful bass on tap. It didn't feel like they needed a subwoofer."

"The drums ... sounded like they were in the room with me. I was very impressed. I didn't want to listen for too long tho. The difference in sq from my more modest system was pretty noticeable."

" [They] were sufficiently impressive enough for me."

"They were pretty amazing"

"I was so impressed with them. If I had the money, I could be talked into them. The cabinets are very inert. It sounded like I was rapping on a solid block when I gave it some knuckles."

"They cost a lot, but you can hear it. Up until then my system was the best I'd heard. Now I know what a really good system can sound like..."

And no, that wasn't some subjective rag writer's comments but those of an AH member who I believe is (or was?) squarely in the die-hard objectivist camp.

I hope he doesn't get upset that I quoted him but it does seem like they left a mark.

 

[AcuDefTechGuy]

Maybe you had them set-up incorrectly??

I am absolutely confident that I set them up correctly for over a year. Many people have come to listen. Some people liked them, some people didn’t like them.

Maybe you’re letting bias get the best of you - hearing things that aren’t there or letting higher volumes dictate quality or falsely equating the +7dB 10kHz accentuated treble as more blackness details.

Here's what one first-time listener had to say about the 802D3:

Have you heard the expression about taking opinions “with a grain of salt” ?

You put way too much stock into people’s opinions.

There’s also this “honeymoon” period that first-time listeners or new owners get when they first buy something new. And then about a year later, they sell their speakers.

You have heard of people selling their B&W Diamond speakers, right?

I sold my 802D2 after raving about them initially. Another forum member here also sold his 802D2 after raving about them initially - when the honeymoon period was over.

The point is that some people like B&W and some people don’t like them.

You love your 805D, the OP loves his 804D, some people don’t.

Nothing else is wrong.

It’s the same way with every brand. There is a speaker for everyone.

 

[Art Vandelay]

Fwiw, I never had a 'honeymoon' with the 800Di's.

It's more of a case of liking them more and more over time. If I upgrade it will be to the 800D3 or possibly the 800D4 when it's released in 2020.

 

[<eargiant]

Maybe you’re letting bias get the best of you -

I see, when someone likes something it's bias (brand/price/etc.) , it can't possibly be because it just sounds better to them. Of course, that doesn't affect you because your ears always agree with the measured graphs. Good thing for you that not many speakers are measured by different testers/facilities in different rooms with different protocols and different published results because then you wouldn't know who or what to believe.

Dennis Murphy recently had his BMRs measured by NRC and the results were different than his own measurements. He contacted them to inquire. If a guy like him needed clarification from NRC as to why the bass was not flat in their measurements do you really think a guy like you or me has all the facts at our disposal to properly interpret a published speaker graph? Yeah, it's fun but I take them with a grain of salt too.

Want to talk about bias? If you had only seen one of the two measurements and you auditioned the BMRs in your home would they have sounded like Dennis' measurements would have led you to believe or would they have sounded like NRCs measurements would led you to believe? I guess it would depend on which measurement you saw right??

hearing things that aren’t there or letting higher volumes dictate quality or falsely equating the +7dB 10kHz accentuated treble as more blackness details.

The "blackness" that's being described has nothing to do with treble. The Polk LSiM 703s also have a spike in the upper regions and they do not come close to the "blackness" in the 800 series. Not even close. BTW, the highs were not as smooth and composed either.

Have you heard the expression about taking opinions “with a grain of salt” ?

Yes, I just did that with the opinion you wrote in the first paragraph of your reply.

You put way too much stock into people’s opinions.

No, just in my own.

But when someone, who based on what I had read here in the past is a vehement objectivist and has the same glowing opinion as I do of a speaker that does not measure objectively "perfect", I take notice.

There’s also this “honeymoon” period that first-time listeners or new owners get when they first buy something new. And then about a year later, they sell their speakers.

You have heard of people selling their B&W Diamond speakers, right?

I sold my 802D2 after raving about them initially. Another forum member here also sold his 802D2 after raving about them initially - when the honeymoon period was over.

Yes, I can agree with you on this point and of course no speaker is perfect, not one (I don't care what a graph may say), a perfect speaker does not exist.

The point is that some people like B&W and some people don’t like them.

You love your 805D, the OP loves his 804D, some people don’t.

Nothing else is wrong.

It’s the same way with every brand. There is a speaker for everyone.

Yes I agree but, there are a lot of speakers raved about here and elsewhere (all price levels) that I found to be meh but I don't knock them or point out there supposed deficiencies every single time someone mentions them on a forum.

For some strange reason you seem hell bent on always jumping in and bashing the 800 series.

It's almost like you're on a personal vendetta against Bowers & Wilkins. By any chance were you ever an authorized dealer?

 

[AcuDefTechGuy]

For some strange reason you seem hell bent on always jumping in and bashing the 800 series.

Well, let’s not get into accusations.

Since I joined this forum many years ago, my views of B&W speakers have been quite consistent.

I was never unequivocally impressed with B&W. I auditioned all the 800s before buying the 802D2. I think the only reason I bought them was because I simply could.

But in general I prefer speakers that have both great sound and measurements.

And in general I agree that people should buy what pleases them.

Everything else is purely academic.

 

[<eargiant]

Well, let’s not get into some personal attacks.

Let’s keep this civilized and fun.

Since I joined this forum many years ago, my views of B&W speakers have been quite consistent.

I was never unequivocally impressed with B&W. I auditioned all the 800s before buying the 802D2. I think the only reason I bought them was because I simply could.

But in general I prefer speakers that have both great sound and measurements.

And in general I agree that people should buy what pleases them.

Apologies if you took that as a personal attack, it was not meant to be one.

Just an observation.

 

[Art Vandelay]

More than most, B&W is a brand that seems to polarise people.

Fwiw, Measurements from B&W's own test facilities look pretty good to me.

 

[William Lemmerhirt]

Tapatalk won’t let me quote but...
That was @Pogre who @eargiant quoted.

 

[adrummingdude]

Hey guys. First time poster, though I'm active on AVSforum.

The discussion going on here is what prompted me to register for this board.

You see, It seems to me that a fair number of posters on this thread are under the impression that a speaker's measurements, especially frequency response, is the be-all-end-all predicter of performance. I believe it is equally important to use your ears, especially in the design stage. In fact, because KEF has been mentioned so many times, and because we all know that they place high importance (as does Harman) on dispersion, its worth mentioning that their design process has actually incorporated more listening tests than they did at one time, yielding what most would consider better products because of it.

Having said that, I would also like to address the B&W Diamond tweeter discussion going on. I have owned this tweeter before, on a pair of 802 D2's, and it's anything but harsh. Weather it has a 10khz bump or not, I often refer to it as the best tweeter I've heard to date, and that includes the beryllium unit in a pair of Focal Sopra 2's I had for a while, and the aluminum dome of the KEF Reference 1 which I recently purchased and am awaiting delivery of.

The thing about 805 D3's not being better than LS50's?...that's just crazy. Yes the LS50 is a great little monitor. I have a pair set up in my office and quite enjoy them. However, having heard the 805 D3 on many occasions I can say that Id absolutely take those over the LS50 in a heartbeat. 805 is a MUCH more open and dynamic speaker. Really, it's not even close. The same thing regarding the chatter of a KEF R700 besting an 804 D3 at high level. Really, I don't know if whoever believes this just loves KEF and hates B&W or what, but it's pretty obvious from some of the comments that there is very little hands on experience involved with the products in question.

I say all that to say this...
Stop staring at graphs and instead go out and listen to the things. Measurements are certainly our friend, but as (even) Dr. Toole has pointed out, it is relatively easy to make a loudspeaker these days with a flat frequency response. That doesn't tell you how the thing is going to sound. Timbre is something that is darn near impossible to measure. At best we can look at the frequency response, decay plot, distortion, dispersion and linearity and using all those metrics have a FAIRLY good idea of how something should sound, but without using our ears we would still be wrong 100% of the time. If you and I were to both sing a pitch perfect note, you would still sound like you and I would sound like me. Loudspeakers are no different.

 

[KenM10759]

Welcome, @adrummingdude to AH. I hang around here too. I completely agree with you. My speakers have to satisfy my ears. For now, they do but there will be something different in the future. It may be KEF Reference or Blade, but just as easily could be a pair I make myself.

I try hard to never bash another's choice (or even opinion) of speakers. My preference is to only make comments about speakers I own now, have owned in the past, or at least gave a "good listen" at my dealer's place. You're right, we all have different ears and taste in what sounds best. I'm now trying to just listen to music, not equipment.

 

[Dennis Murphy]

Hey guys. First time poster, though I'm active on AVSforum.

The discussion going on here is what prompted me to register for this board.

You see, It seems to me that a fair number of posters on this thread are under the impression that a speaker's measurements, especially frequency response, is the be-all-end-all predicter of performance. I believe it is equally important to use your ears, especially in the design stage. In fact, because KEF has been mentioned so many times, and because we all know that they place high importance (as does Harman) on dispersion, its worth mentioning that their design process has actually incorporated more listening tests than they did at one time, yielding what most would consider better products because of it.

Having said that, I would also like to address the B&W Diamond tweeter discussion going on. I have owned this tweeter before, on a pair of 802 D2's, and it's anything but harsh. Weather it has a 10khz bump or not, I often refer to it as the best tweeter I've heard to date, and that includes the beryllium unit in a pair of Focal Sopra 2's I had for a while, and the aluminum dome of the KEF Reference 1 which I recently purchased and am awaiting delivery of.

The thing about 805 D3's not being better than LS50's?...that's just crazy. Yes the LS50 is a great little monitor. I have a pair set up in my office and quite enjoy them. However, having heard the 805 D3 on many occasions I can say that Id absolutely take those over the LS50 in a heartbeat. 805 is a MUCH more open and dynamic speaker. Really, it's not even close. The same thing regarding the chatter of a KEF R700 besting an 804 D3 at high level. Really, I don't know if whoever believes this just loves KEF and hates B&W or what, but it's pretty obvious from some of the comments that there is very little hands on experience involved with the products in question.

I say all that to say this...
Stop staring at graphs and instead go out and listen to the things. Measurements are certainly our friend, but as (even) Dr. Toole has pointed out, it is relatively easy to make a loudspeaker these days with a flat frequency response. That doesn't tell you how the thing is going to sound. Timbre is something that is darn near impossible to measure. At best we can look at the frequency response, decay plot, distortion, dispersion and linearity and using all those metrics have a FAIRLY good idea of how something should sound, but without using our ears we would still be wrong 100% of the time. If you and I were to both sing a pitch perfect note, you would still sound like you and I would sound like me. Loudspeakers are no different.

I've always understood Floyd to say that a speaker that measures well on-axis and avoids serious peaks and dips off axis (other than the normal tweeter roll off at higher frequencies) will be preferred by most listeners. It's really not all that easy to achieve that kind of response, and a number of B & W speakers fall pretty short. That said, I personally have always liked the original 800 series and haven't had an opportunity to hear the latest ones.

 

[adrummingdude]

I've always understood Floyd to say that a speaker that measures well on-axis and avoids serious peaks and dips off axis (other than the normal tweeter roll off at higher frequencies) will be preferred by most listeners. It's really not all that easy to achieve that kind of response, and a number of B & W speakers fall pretty short. That said, I personally have always liked the original 800 series and haven't had an opportunity to hear the latest ones.

Yes, that is what I have understood from Floyd as well. He places a huge emphasis on off axis frequency response with Harman's Spinorama chamber measurements.

Where I think he is maybe a bit off though is that "preferred by most listeners" shouldn't really hold much water. Most of those listeners probably also preferred no doc adjustable rate loans, popcorn ceiling, and Nickelback.

The point being that Harman under Toole's guidance, as I understand it, have set up their test to use a mostly untrained listeners who they then "prep" to listen for certain characteristics...that just happen to coincide with a Revel speaker's strengths. Conversely, if they were using mostly experienced yet unprepped enthusiasts like the gents found on boards like this I would be much more willing to accept their results. Finally, I have yet to hear a loudspeaker that plays EVERYTHING well. So, even the tailored B&W response will sound much better playing certain tracks that stack nicely to that response than will say a more neutral speaker, and double blind or not, somewhere at Harman is a master test doer dude in charge of curating a playlist. Just sayin.

 

[Dennis Murphy]

Yes, that is what I have understood from Floyd as well. He places a huge emphasis on off axis frequency response with Harman's Spinorama chamber measurements.

Where I think he is maybe a bit off though is that "preferred by most listeners" shouldn't really hold much water. Most of those listeners probably also preferred no doc adjustable rate loans, popcorn ceiling, and Nickelback.

The point being that Harman under Toole's guidance, as I understand it, have set up their test to use a mostly untrained listeners who they then "prep" to listen for certain characteristics...that just happen to coincide with a Revel speaker's strengths. Conversely, if they were using mostly experienced yet unprepped enthusiasts like the gents found on boards like this I would be much more willing to accept their results. Finally, I have yet to hear a loudspeaker that plays EVERYTHING well. So, even the tailored B&W response will sound much better playing certain tracks that stack nicely to that response than will say a more neutral speaker, and double blind or not, somewhere at Harman is a master test doer dude in charge of curating a playlist. Just sayin.

I believe the results were the same for experienced listeners. But I don't know the ins and outs of the training sessions and the choice of music. My experience designing speakers doesn't totally jibe with Floyd's extreme emphasis on far-off-axis response. I do believe in wide dispersion, but I'm not convinced that a peak or a dip at 60 degrees off axis is going to be much of a problem.

 

[adrummingdude]

I believe the results were the same for experienced listeners. But I don't know the ins and outs of the training sessions and the choice of music. My experience designing speakers doesn't totally jibe with Floyd's extreme emphasis on far-off-axis response. I do believe in wide dispersion, but I'm not convinced that a peak or a dip at 60 degrees off axis is going to be much of a problem.

A peak or dip at 60 degrees off axis doesn't matter much to me either, but I think where Floyd may go with that is that the first reflection points will now be inconsistent with the direct on-axis sound. Meh...whatever.

Interestingly, I recall reading or hearing that the untrained listeners in the Harman tests were more reliable in identifying particular speakers than were hifi sales staff or professional reviewers. I don't know exactly what that information is worth, but it is interesting.

I don't have any experience designing speakers, but I will say that speakers with a tailored frequency response like the B&W's being discussed in this thread I have very much enjoyed. Not all of them, but a good many anyways.

To use a car analogy...because I'm sure THAT has never been done before on a hifi forum..., its like having a $75k Lexus and a $75k Jaguar. Unquestionably the Lexus is the better car. It would be the one with the better off axis frequency response, more reliable, better value, preferred by many, many more "listeners" and they have the sales numbers to prove it....but, the Jag is sexy, no? Yeah, it might leave you broken down by the side of the road and good luck finding a local stealership to service the thing, but man, that sumabich is fast and sounds great! There's merit to that, even if it is the "wrong" choice.

 

[PENG]

I thought the AH article (by Dr. F.Toole and Gene) linked below summed it up quite well, for me anyway.

https://www.audioholics.com/loudspeaker-design/measure-loudspeaker-performance

Follow is the part that pretty much confirmed what Dennis said he understood, he must have good memory. The fact that he is a speaker design specialist surely helps though.

In a nutshell, if a speaker exhibits flat and linear on-axis frequency response with consistent off-axis performance to preserve critical early reflections, then the speaker will score very highly in blind listening tests and also provide more consistent performance from room to room. The speaker must also be free of pesky resonances and the transition between drivers must be as seamless as possible, hence why waveguides are an essential part in most Harman loudspeaker products. The waveguide not only increases sensitivity at lower frequencies, but allows the high frequency driver (aka. the tweeter) to match directivity of the midrange at the critical crossover region.

Another one I think is a good read is:

https://www.harman.com/sites/default/files/AudioScience_0.pdf

On page 10 he talked about "Blind vs Sighted test". We know how controversial that one is, even for the simplistic power amplifiers, imagine speakers.

I respect the gentleman a lot and don't doubt his knowledge and experience, but I do take some of the things he stated in some articles and videos with a grain of salt.

I like the original 802D, and even the affordable 804S a lot, but based on distant memories I would prefer the KEF Blades.

It is very subjective thing to say one sounds better than the other to begin with. To compare speakers in different rooms at different time makes it much more than just a "subjective" thing. So I believe it is entirely possible that if I were to compare the Blade and any of the 802 or 800 diamond versions on a turntable, I may still prefer the B&W, or the Blade. You wonder why I do rely a lot on, and trust measurements, while I fully recognize the importance of using my ears and will not buy any expensive speakers without auditioning them to the nth degree.

 

[adrummingdude]

It is very subjective thing to say one sounds better than the other to begin with. To compare speakers in different rooms at different time makes it much more than just a "subjective" thing. So I believe it is entirely possible that if I were to compare the Blade and any of the 802 or 800 diamond versions on a turntable, I may still prefer the B&W, or the Blade. You wonder why I do rely a lot on, and trust measurements, while I fully recognize the importance of using my ears and will not buy any expensive speakers without auditioning them to the nth degree.

Yup.

Its also important realize that we are discussing very nice speakers here. When you get over $10k per pair, there are a lot fewer compromises happening. They all have good drivers, good cabinets, good crossovers, and have been completely thought out by men who know what they're doing. That the different models from different manufacturers don't sound alike is IMO secondary to them all sounding lovely, because at this level they all do.

 

[Dennis Murphy]

A peak or dip at 60 degrees off axis doesn't matter much to me either, but I think where Floyd may go with that is that the first reflection points will now be inconsistent with the direct on-axis sound. Meh...whatever.

Interestingly, I recall reading or hearing that the untrained listeners in the Harman tests were more reliable in identifying particular speakers than were hifi sales staff or professional reviewers. I don't know exactly what that information is worth, but it is interesting.

I don't have any experience designing speakers, but I will say that speakers with a tailored frequency response like the B&W's being discussed in this thread I have very much enjoyed. Not all of them, but a good many anyways.

To use a car analogy...because I'm sure THAT has never been done before on a hifi forum..., its like having a $75k Lexus and a $75k Jaguar. Unquestionably the Lexus is the better car. It would be the one with the better off axis frequency response, more reliable, better value, preferred by many, many more "listeners" and they have the sales numbers to prove it....but, the Jag is sexy, no? Yeah, it might leave you broken down by the side of the road and good luck finding a local stealership to service the thing, but man, that sumabich is fast and sounds great! There's merit to that, even if it is the "wrong" choice.

well, we agree on cars at least. But I dob't think there's any standard definition of a "high fidelity" car, and there is a good definition of a "high fidelity" loudspeaker. So I don't think the analogy holds up very well. As designer I'm trying to achieve the most accurate reproduction of the original event as possible. I'll never achieve that, of course, and in some cases there isn't really any event to replicate, but I definitely won't tailor the response to produce a "house" sound, or deliberately introduce a dip in the mid treble to make certain recordings more palatable.

 

[<eargiant]

Hey guys. First time poster, though I'm active on AVSforum.

The discussion going on here is what prompted me to register for this board.

You see, It seems to me that a fair number of posters on this thread are under the impression that a speaker's measurements, especially frequency response, is the be-all-end-all predicter of performance. I believe it is equally important to use your ears, especially in the design stage. In fact, because KEF has been mentioned so many times, and because we all know that they place high importance (as does Harman) on dispersion, its worth mentioning that their design process has actually incorporated more listening tests than they did at one time, yielding what most would consider better products because of it.

Agreed

Having said that, I would also like to address the B&W Diamond tweeter discussion going on. I have owned this tweeter before, on a pair of 802 D2's, and it's anything but harsh. Weather it has a 10khz bump or not, I often refer to it as the best tweeter I've heard to date, and that includes the beryllium unit in a pair of Focal Sopra 2's I had for a while, and the aluminum dome of the KEF Reference 1 which I recently purchased and am awaiting delivery of.

Agreed- based on hands on ownership experience.

The thing about 805 D3's not being better than LS50's?...that's just crazy. Yes the LS50 is a great little monitor. I have a pair set up in my office and quite enjoy them. However, having heard the 805 D3 on many occasions I can say that Id absolutely take those over the LS50 in a heartbeat. 805 is a MUCH more open and dynamic speaker. Really, it's not even close. The same thing regarding the chatter of a KEF R700 besting an 804 D3 at high level. Really, I don't know if whoever believes this just loves KEF and hates B&W or what, but it's pretty obvious from some of the comments that there is very little hands on experience involved with the products in question.

Agreed-based on hands on ownership experience of both at the same time.

I say all that to say this...
Stop staring at graphs and instead go out and listen to the things. Measurements are certainly our friend, but as (even) Dr. Toole has pointed out, it is relatively easy to make a loudspeaker these days with a flat frequency response. That doesn't tell you how the thing is going to sound. Timbre is something that is darn near impossible to measure. At best we can look at the frequency response, decay plot, distortion, dispersion and linearity and using all those metrics have a FAIRLY good idea of how something should sound, but without using our ears we would still be wrong 100% of the time. If you and I were to both sing a pitch perfect note, you would still sound like you and I would sound like me. Loudspeakers are no different.

Save your breath adrummingdude, it's no use.

Thanks for the heads-up, I'm going to check out AVS Forums...

 

[shadyJ]

Yes, that is what I have understood from Floyd as well. He places a huge emphasis on off axis frequency response with Harman's Spinorama chamber measurements.

Where I think he is maybe a bit off though is that "preferred by most listeners" shouldn't really hold much water. Most of those listeners probably also preferred no doc adjustable rate loans, popcorn ceiling, and Nickelback.

The point being that Harman under Toole's guidance, as I understand it, have set up their test to use a mostly untrained listeners who they then "prep" to listen for certain characteristics...that just happen to coincide with a Revel speaker's strengths. Conversely, if they were using mostly experienced yet unprepped enthusiasts like the gents found on boards like this I would be much more willing to accept their results. Finally, I have yet to hear a loudspeaker that plays EVERYTHING well. So, even the tailored B&W response will sound much better playing certain tracks that stack nicely to that response than will say a more neutral speaker, and double blind or not, somewhere at Harman is a master test doer dude in charge of curating a playlist. Just sayin.

Harman didn't design listener preference tests that suited the strengths of their speakers, they designed speakers that conformed to what they found in listener preference tests. And their tests used both trained and untrained listeners. And no they did not compile a playlist that played to the strengths of their speakers.

The research done by Toole and his colleagues are the most extensive testing performed that gauges what is commonly thought to be a "good sound" from loudspeakers, and that happens to correspond extremely well to an actual accurate (neutral) response. This was established over years of testing by a team of very intelligent Ph.Ds and a relatively large sample size and variety of listeners. I don't even know of any other manufacturer or institution that has used real science to establish what sounds good to human hearing. I imagine that it has been done, but not to the extant that Harman has done- and I doubt that such research would have significantly differed from the conclusions reached by Harman's research.

To blithely dismiss that extraordinary body of research with a handful of ill-informed critiques is fantastically ignorant.

 

[KEW]

Harman didn't design listener preference tests that suited the strengths of their speakers, they designed speakers that conformed to what they found in listener preference tests. And their tests used both trained and untrained listeners. And no they did not compile a playlist that played to the strengths of their speakers.

The research done by Toole and his colleagues are the most extensive testing performed that gauges what is commonly thought to be a "good sound" from loudspeakers, and that happens to correspond extremely well to an actual accurate (neutral) response. This was established over years of testing by a team of very intelligent Ph.Ds and a relatively large sample size and variety of listeners. I don't even know of any other manufacturer or institution that has used real science to establish what sounds good to human hearing. I imagine that it has been done, but not to the extant that Harman has done- and I doubt that such research would have significantly differed from the conclusions reached by Harman's research.

To blithely dismiss that extraordinary body of research with a handful of ill-informed critiques is fantastically ignorant.

I had the same concern about how the Harman research was characterized in this thread.

It makes total sense for a speaker manufacture willing to devote the resources to attempt to determine what characteristics of speakers result in them being favored speakers. That is, simply good market research.
If they designed the testing simply to favor their own speakers, that would be counter productive as they have no idea how well their speaker would be received, they are simply patting each other on the shoulder and self-congratulations. I have met many intelligent people who are driven to do research because they desire answers to questions that do not have obvious answers. I am convinced that Floyd is one of them. For these guys, the design of their experiments to provide the best answers they can is the only objective.