Comparison of B & W 800 D and B & W 800 D3

[TLS Guy]

B & W 800 D

The B & W 800 D3

My friend and former colleague in arms so to speak had been becoming increasingly dissatisfied with his B & W 800 Ds.

Around the Christmas period I was demonstrating my new Intel NUC i5. Phil has always liked the speakers here at Eagan since I built them in 1989. Our Dermatologist at the clinic and an audio enthusiast was particularly keen on these speakers. I built them as my location monitors, which is why they are on wheels.

Anyhow Phil seemed to be paying more attention to the speakers than the NUC and frequently complimenting the sound of the speakers.

Right after Christmas he placed an order for a pair B & W 800 D3s, without audition. Remember these are $30,000 per pair. The dealer allowed him $9000 for the 800 Ds and his B & W sub.

The speakers were shipped from the UK and arrived in Minneapolis at the end of last week and were delivered today.

The major problem with the 200 Ds was that there was a little sibilance to speech, the violins were not as smooth as they should be and the sound stage lacked depth.

In addition the bass was a shade plumy with noticeable port kick in. The speakers did benifit from a sub but integration was always problematic.

Last night I did a quick capture of the on axis frequency response and the impulse response of the B & W 800 Ds

Now this shows how little it takes to muck up a speaker. That hump from the kevlar cone break up mode between 3 KHz and 4.5 KHz is all it takes to upset the apple cart so to speak. The warmth is caused by the broad peak from 60 to 150 Hz. Port kick in can be seen below 30 Hz.

Immediate listening tests of the 800 D3 speakers were very encouraging. First up was Haydn's Mass in Time of War from this years Prom with Kings College and the Orchestra of the Age of Enlightenment under Stephen Cleobury.

Right away the bass was tight and perfectly balanced without bloom. The tymps which feature prominently were sharp and crisp. There was no overhang or boom. The strings and voices were smooth and the sound stage spectacular. The sound stage was wide and deep with much os the RAH acoustic preserved.

After that listening session I obtained this result under identical conditions to the test on the 800 Ds

The real problem area at 3 to 4 KHz is gone. There is a very slight rise from 60 to 100 Hz, but this may well be an artifact, and in any case if real would be inaudible.

The bass extends to 20 Hz and so this is a true full range speaker. The port cut in is far less pronounced.

The impulse response is improved as is usually the case as things improve.

Extended listening test today confirm that this is a first class speaker. It has quite the tightest bass of any porter speaker I have heard.

Listening tests with large choral works and big powerful organs confirmed that no sub is required for these speakers and in fact would be a detriment.

The sound is incredibly similar to the sound of my reference speakers. I would say that mine have perhaps marginally more visceral impact, but it is hard to be certain as they are in different rooms.

These speakers being passive, there is no way to electronically blend in the LFE channel. My speakers being triamped the LFE can be correctly mixed and blended in.

In summary I have to say that these 800 D3 speakers are worthy to be the flagship of the new B & W range. These are first class speakers, and ones that I am certain I could live with for year after year. I could not say the same about its predecessors.

Well done B & W. you have put right what has needed putting right for some time.

 

[BoredSysAdmin]

One little issue, Mark..
800D measurement is nice 50db scale, while D3 is graph scale is 100db....

 

[TLS Guy]

One little issue, Mark..
800D measurement is nice 50db scale, while D3 is graph scale is 100db....

Well that is what happens when there is a ton of people in the room and it is two different days. Sorry about that, but you can still see the improvement, especially in the 3 to 4K Hz area. That was the major problem by far.

 

[Dennis Murphy]

Well that is what happens when there is a ton of people in the room and it is two different days. Sorry about that, but you can still see the improvement, especially in the 3 to 4K Hz area. That was the major problem by far.

A couple of comments about the measurements (in addition to the scale issue), which I assume were done with the Dayton OmniMic. First, it looks like you've gone with the default smoothing function (1/6 Octave), which obscures a lot of information. 1/24 would be more informative, particularly in seeing the full contours of the 4-5 kHz peak in the Ds. Second, it appears that your omnimic software, like every other sample I've used or seen used, suffers from a inaccurate roll-off above 10 kHz. I'm not sure what causes this problem, but I would bet my government pension that the B&W's are actually flat between 10k and 20k, or even rising.

 

[lovinthehd]

A couple of comments about the measurements (in addition to the scale issue), which I assume were done with the Dayton OmniMic. First, it looks like you've gone with the default smoothing function (1/6 Octave), which obscures a lot of information. 1/24 would be more informative, particularly in seeing the full contours of the 4-5 kHz peak in the Ds. Second, it appears that your omnimic software, like every other sample I've used or seen used, suffers from a inaccurate roll-off above 10 kHz. I'm not sure what causes this problem, but I would bet my government pension that the B&W's are actually flat between 10k and 20k, or even rising.

Tell us about this government pension... ;

 

[TLS Guy]

A couple of comments about the measurements (in addition to the scale issue), which I assume were done with the Dayton OmniMic. First, it looks like you've gone with the default smoothing function (1/6 Octave), which obscures a lot of information. 1/24 would be more informative, particularly in seeing the full contours of the 4-5 kHz peak in the Ds. Second, it appears that your omnimic software, like every other sample I've used or seen used, suffers from a inaccurate roll-off above 10 kHz. I'm not sure what causes this problem, but I would bet my government pension that the B&W's are actually flat between 10k and 20k, or even rising.

I think the mic must roll off. Every speaker I have tested with that rig rolls off at the top end like that, soI agree it is a feature of the program.

 

[PENG]

My REW/Umik-1 miniDSP rolls off too in that range. I thought, as Audyssey has been saying, it is like that in live music (concert halls).

 

[Swerd]

… That hump from the kevlar cone [in the 800D] break up mode between 3 KHz and 4.5 KHz is all it takes to upset the apple cart so to speak. The warmth is caused by the broad peak from 60 to 150 Hz. Port kick in can be seen below 30 Hz.

Immediate listening tests of the 800 D3 speakers were very encouraging. The real problem area at 3 to 4 KHz is gone. There is a very slight rise from 60 to 100 Hz, but this may well be an artifact, and in any case if real would be inaudible.

Thanks for posting the response graphs, the photos, and your comments.

I find it amazing that B&W persisted with those yellow Kevlar mids for all these years. Even in the 800 series speakers, the crossover was so high that it allowed the breakup mode to bleed through. In the CM and 600 series it sounded worse. The new 800 D3 speakers replaced the mid with a new cone material. Is it metal? But the audible problem with the old Kevlar driver's could have been more easily fixed anytime by using a lower crossover frequency, closer to 2 KHz. I never understood why B&W wouldn't do that.

For 30 k$/pair B&W has a lot of competition. I'll probably never hear them, but I do wonder how the new 800 D3 would compare to the much lower priced Philharmonic 3 or the Salk SoundScape 8?

 

[Dennis Murphy]

My REW/Umik-1 miniDSP rolls off too in that range. I thought, as Audyssey has been saying, it is like that in live music (concert halls).

But that makes it very difficult to design speakers that start out flat. High frequencies do tend to get absorbed in concert venues and in certain listening rooms, but the source of the high frequencies isn't rolled off. Otherwise, the OmniMic is a very flexible, easy to use tool. But I've had to give up on it for designing speakers.

 

[Irvrobinson]

A couple of comments about the measurements (in addition to the scale issue), which I assume were done with the Dayton OmniMic. First, it looks like you've gone with the default smoothing function (1/6 Octave), which obscures a lot of information. 1/24 would be more informative, particularly in seeing the full contours of the 4-5 kHz peak in the Ds. Second, it appears that your omnimic software, like every other sample I've used or seen used, suffers from a inaccurate roll-off above 10 kHz. I'm not sure what causes this problem, but I would bet my government pension that the B&W's are actually flat between 10k and 20k, or even rising.

I've found that the OmniMicII version seems to at least mitigate the high-end roll-off problem. I've owned both versions, and noticed the difference (or at least thought I did).

I agree about using 1/6 octave smoothing. I find it pretty much useless, of course I use the OmniMic mostly for bass tuning, and figuring out issues in the 100Hz-2000Hz range. For anything up to about 1KHz I think no smoothing is best.

 

[Irvrobinson]

In summary I have to say that these 800 D3 speakers are worthy to be the flagship of the new B & W range. These are first class speakers, and ones that I am certain I could live with for year after year. I could not say the same about its predecessors.

Well done B & W. you have put right what has needed putting right for some time.

Well, it's about time that B&W built a speaker that doesn't appear to be targeted to be a showroom traffic attractor. Too bad that at $30K I doubt I'd ever consider it, even with a 15% B&M discount. I'm especially surprised at that price considering the relative strength of the US dollar and the weakness of the Pound. Pretty bold.

 

[PENG]

But that makes it very difficult to design speakers that start out flat. High frequencies do tend to get absorbed in concert venues and in certain listening rooms, but the source of the high frequencies isn't rolled off. Otherwise, the OmniMic is a very flexible, easy to use tool. But I've had to give up on it for designing speakers.

The miniDSP mic is not too bad because if I position the mic 12" to 1 meter from the speaker, the graph does not show much of a roll off until around 17.5kHz, from 11 to 12', roll off starts at around 12-13kHz.

https://www.minidsp.com/products/acoustic-measurement/umik-1

 

[PENG]

I've found that the OmniMicII version seems to at least mitigate the high-end roll-off problem. I've owned both versions, and noticed the difference (or at least thought I did).

I agree about using 1/6 octave smoothing. I find it pretty much useless, of course I use the OmniMic mostly for bass tuning, and figuring out issues in the 100Hz-2000Hz range. For anything up to about 1KHz I think no smoothing is best.

I think the similarly prices KEF Blade would sound better but the B&W has the much better look, but then that is highly subjective.

 

[TLS Guy]

A couple of comments about the measurements (in addition to the scale issue), which I assume were done with the Dayton OmniMic. First, it looks like you've gone with the default smoothing function (1/6 Octave), which obscures a lot of information. 1/24 would be more informative, particularly in seeing the full contours of the 4-5 kHz peak in the Ds. Second, it appears that your omnimic software, like every other sample I've used or seen used, suffers from a inaccurate roll-off above 10 kHz. I'm not sure what causes this problem, but I would bet my government pension that the B&W's are actually flat between 10k and 20k, or even rising.

Dennis, I returned today, and got date with some peace and quiet, and did more listening tests.

The B & W 800 D is a very nice speaker. It is not forward like the older versions. The sound stage is nice and recessed.

Here is the FR on and off axis. I went all the way to 90 degrees, which is not a total disaster.

Axis Blue
15 degrees off axis Yellow
45 degrees off axis Red
60 degrees off axis Green
90 degrees off axis Black

All measurements 1 meter on tweeter axis.

I don't really care about the roll off real or not. The extreme high end has little to do with how a good speaker really sounds, unless the high end rises.

These speakers are just a fraction retiring which I like. The older versions of the 800 series I think were aimed to please the pop enthusiast. To me these speakers seem squarely voiced at the classical music listener, which is as it should be, as that is really the bread and butter of the high end speaker market. It really is of no consequence what rock and pop sound like, since no one knows what it is supposed to sound like in the first place!

I think the really good off axis response is why these speakers have such a good sound stage.

I did take a quick snap shot if my Eagan speakers which were the immediate precipitation to Phil forking out 30K on new speakers. I have never bothered to measure these speakers. They were built in 1989. I used Bullock and White's software on an Apple IIe, for enclosures and crossovers. I did the critical voicing entirely by ear as so many did back then.

It is an unusual design. It is basically a 3.5 way. Tweeter is Scanspeak with two Dynaudio large VC drivers. The crossover is 3 KHz to the tweeter. The two Dynaudios each have their own sealed enclosure, so one can not drive the other. F3 is 90 Hz, with electronic crossover at 90 Hz second order for composite fourth order. The lower driver provides BSC. The low end is two Morel 9" subwoofer drivers, in coupled cavity isobarik configuration. Calculated f3 points are 90 Hz and 27 Hz. With this bandwidth the Q of the bass system is 0.7. The sub drivers are driven to 120 Hz so the high end roll off of the bass system is essentially acoustic.

The whole purpose of the design was to provide for location monitors for my radio broadcasts and recordings, that was not massive and had a very respectable low end response.

Musicians really used to like these speakers on session playback. My wife loves these speakers, but not the size!

 

[Dennis Murphy]

The on axis response is pretty good, although not really top notch. I can see why you think they're a little retiring. That's a matter of taste, I guess, but I would still prefer to hear a recording exactly the way it left the studio. As for the top octave, the problem with the OmiMic is that the response may very well be rising. Virtually every speaker I've measured with OmniMic droops in the top octave, including some that are actually voiced quite hot in that region. Of course, all of this will be more of an issue to younger listeners than fossils like me.

 

[TLS Guy]

The on axis response is pretty good, although not really top notch. I can see why you think they're a little retiring. That's a matter of taste, I guess, but I would still prefer to hear a recording exactly the way it left the studio. As for the top octave, the problem with the OmiMic is that the response may very well be rising. Virtually every speaker I've measured with OmniMic droops in the top octave, including some that are actually voiced quite hot in that region. Of course, all of this will be more of an issue to younger listeners than fossils like me.

I think that both the axis and off axis responses are good. The response is not everything. The midrange driver is wide band. I believe firmly that crossovers are a necessary evil. They play fast and loose with time, and cause horrible measurable problems. We just don't usually measure them. I personally believe there is strong merit to keeping crossovers out of the area where the human ear is most sensitive. I think B & W are correct in that long held opinion. I think the way to better speakers is through the use of wide band drivers.

In addition I'm convinced that passive components get progressively problematic the lower the crossover point. I think B & W should make an active version of that speaker. I think exotic expensive speakers should be active now, universally.

Phil is driving those speaker with 450 watts per channel Mac amps. On the big forces things sound very slightly stressed the way mine do not with the avoidance of passive components in the lower octaves. I would bet an active version of that speaker would be a worthwhile improvement.

As to measuring mics, I don't think you will get a mic that does not roll off in the top end unless the mic is very expensive. I have a good mic selection. The only mics that really extend flat out to 20 KHz are the very expensive condenser Neumann's that cost in the thousands.

I think putting crossovers above 10 KHz in speakers is goofy. I think with tweeters from the likes of SEAS and Scanspeak you can rely on the on the manufacturers published acoustic response graphs. In any event you can tame a rising response, but pulling a drooping response up with a passive crossover is usually not feasible, unless here is a rising response before the falling response.

As to what the sound was like leaving the studio, we don't know unless we are the recording and post production engineer. I have made hundreds of live recordings over past years, and quite a few under studio type conditions. I can tell you this, we all mix to our speakers in the original monitoring room and the speakers in post production. Inevitably the sound stage presented by these speakers will have a huge impact on the "sound that left the studio."

The BBC engineers in my view have always set the highest standards of the art. It is now nice to be able to get really nice examples of their work from BBC UK via VPN tunnel.

Most of the sources I used to evaluate those B & W speakers were from CDs I made from BBC Prom Broadcasts and Choral Evensong.

I can tell you they did a very good job of these. In the Prom discs the sound stage form the RAH was laid out in a huge deep arc, and was very realistic, just like I experience with my speakers.

I personally have zero tolerance of speakers that keep the sound stage confined between the speakers, or worse in front of them. The 800 Ds never gave me that deep sound stage, but the D3s do.

 

[TLS Guy]

Thanks for posting the response graphs, the photos, and your comments.

I find it amazing that B&W persisted with those yellow Kevlar mids for all these years. Even in the 800 series speakers, the crossover was so high that it allowed the breakup mode to bleed through. In the CM and 600 series it sounded worse. The new 800 D3 speakers replaced the mid with a new cone material. Is it metal? But the audible problem with the old Kevlar driver's could have been more easily fixed anytime by using a lower crossover frequency, closer to 2 KHz. I never understood why B&W wouldn't do that.

For 30 k$/pair B&W has a lot of competition. I'll probably never hear them, but I do wonder how the new 800 D3 would compare to the much lower priced Philharmonic 3 or the Salk SoundScape 8?

The midrange cone is a weave, I suspect of carbon fiber, with a metallic glue spray. I'm pretty sure they are sprayed with something that contains metal. From my listening and measurements I don't think that cone is breaking up at least irregularly in its operating range. A speaker can have cone break up in its operating range and still sound and measure well. It is how it breaks up that is the issue. That was always Ted Jordan's point, and I believe him to be correct, that you need to accept cone break up, but it needs to be predictable and controlled. After all, soft dome tweeters sound very good. They are my preferred tweeters, yet they all break up!

If B & W had lowered the crossover to 2 KHz that would have put it far too close to the 350 Hz crossover. 3 KHz they could have got away with. I like to cross to tweeters 3 KHz and above if I possibly can. Unfortunately that is so often not possible.

Finally at least for classical listening I believe a speaker that encompasses the whole musical spectrum without resort to a sub is the preferred approach. The 800 D3 is a full range speaker with correctly balanced lower octaves.

 

[highfigh]

It really is of no consequence what rock and pop sound like, since no one knows what it is supposed to sound like in the first place!

Once again, you're using musical snobbery to discount music that did more to advance recording technology than any other. The goal of some producers of Rock and Pop is to provide a recording that accurately reflects the performance and for many, the sound quality is very important. Then, there are the others who, admittedly, don't seem to know or care that it's actually possible to make an accurate recording. It's not all electric instruments and screeching vocals.

I'm not sure if you know or care, but that attitude is very insulting. If someone who prefers styles other than Classical said/posted that Classical Music is stodgy, the standard repertoire is old & doesn't change and it's stuck in the past, I'm pretty sure you would be enraged but those views, which are just opinions, are based on ignorance. You're saying that there's no importance in Rock & Pop sounding good but that's just wrong and it's based on the same kind of ignorance. Many people enjoy the music they heard when they were growing up but keeping an open mind isn't a bad thing. OTOH, every generation hates the music of the next, so....

I don't like a lot of what has been trotted out as "Pop" music in the last 35 years, so I can relate to the idea that it's not good, but almost all forms of music contribute to advances in recording techniques. The death of dynamic range and sound quality aren't the fault of the musical styles, they're the result of the bad technologies associated with small, battery-powered hand held devices and ear buds.

However, to the point about the sound from the 800D, if your opinion of Rock & Pop is partially due to the sound quality from these speakers, I think I can understand it- I was in the control room of a studio when a friend was recording some tracks for one of his CDs and he was using one of my guitar amps, which I have owned for over 30 years and I'm familiar with its sound- I have set the controls in just about every position possible and what came from those speakers DID NOT sound like my amp.

 

[PENG]

I don't find that insulting at all, generalizing may be, but not insulting.

 

[shadyJ]

Once again, you're using musical snobbery to discount music that did more to advance recording technology than any other. The goal of some producers of Rock and Pop is to provide a recording that accurately reflects the performance and for many, the sound quality is very important. Then, there are the others who, admittedly, don't seem to know or care that it's actually possible to make an accurate recording. It's not all electric instruments and screeching vocals.

I'm not sure if you know or care, but that attitude is very insulting. If someone who prefers styles other than Classical said/posted that Classical Music is stodgy, the standard repertoire is old & doesn't change and it's stuck in the past, I'm pretty sure you would be enraged but those views, which are just opinions, are based on ignorance. You're saying that there's no importance in Rock & Pop sounding good but that's just wrong and it's based on the same kind of ignorance. Many people enjoy the music they heard when they were growing up but keeping an open mind isn't a bad thing. OTOH, every generation hates the music of the next, so....

I don't like a lot of what has been trotted out as "Pop" music in the last 35 years, so I can relate to the idea that it's not good, but almost all forms of music contribute to advances in recording techniques. The death of dynamic range and sound quality aren't the fault of the musical styles, they're the result of the bad technologies associated with small, battery-powered hand held devices and ear buds.

However, to the point about the sound from the 800D, if your opinion of Rock & Pop is partially due to the sound quality from these speakers, I think I can understand it- I was in the control room of a studio when a friend was recording some tracks for one of his CDs and he was using one of my guitar amps, which I have owned for over 30 years and I'm familiar with its sound- I have set the controls in just about every position possible and what came from those speakers DID NOT sound like my amp.

I think TLS Guy's points do have some merit, although he does deliver them with a dose of audio snobbery. Many classical recordings do have a much wider dynamic range than most pop music recordings, but there is more to it than that. Many classical recordings also have a reference point for their sound, ie they are trying to capture the sound of a certain venue. They are trying to make the music sound lifelike in the sense that is what you would hear in good seats in a concert hall. Pop music, on the other hand, isn't recorded to sound like anything that would be realistically be heard in real life. The 'scene' created by most pop music occurs in a purely psychic space. Mostly it is just recorded so all the notes, lyrics, and sound are heard clearly, and is not trying to recreate a real life acoustic space. As such, there is no reference point for the way it is supposed to sound, at least if you are not the recording engineer.