DYNAUDIO C4 VS B&W 800 D3

[TLS Guy]

Huge impedance swings indicate poorly paired drivers and crossovers.

The measurements show the Revel playing lower while B&W has more amplitude but rolls off sharply below 30 hz, right between those 'off the chart' impedance swings.



I just take issue with such an expensive speaker having any measurable defects built in, especially for the sake of 'plastic styling panels.'



Can you share those measurements? I can't find any!



Inert cabinet construction is pretty well understood, I can't actually think of any models that do suffer from poor cabinet construction.

Tweeter pod, you mean the B&W baffle-less tweeter? The baffle serves as the launch pad for the sound wave. Without it, the directivity of the high frequencies is even more narrow, making for the timbral shift that occurs when one moves off axis.

I think B&W's sounds great on axis. I just don't think they deserve so much credit when you consider the price, they don't sound the same to every seat. But it seems their function following form design is too attractive for some, to care.

The D3s do not have a Kevlar midrange driver. It is a carbon fiber weave. It sounds significantly different to the old Kevlar driver.

In addition the new D3 does sound very similar right across the room, the old Kevlar driver did not.

What I found was a slight dip at 5 KHz, and not near as high a peak at 10 HHz. My measurements agree with the sonic footprint of the speaker very well.

The Ds were forward, the D3s are retiring and a very different speaker.

After a listening session at Eagan to my rig there, Phil realized it was time for a change. He has always loved those speakers. He is delighted with his 800 D3 and they sound much more like my designs.

 

[TheWarrior]

Can you elaborate on this? I would think "poorly paired drivers and crossovers" would manifest as poor frequency response, driver directivity mismatch, out-of-band resonances insufficiently filtered, tweeters operating too low and thus limiting power handling, etc.

According to the graphs above by ADTG, not a single model can maintain even a +/-3dB spec above 200 hz. I agree with you that below 200 hz shows the only significant changes, as all 3 models measure almost identically above 200 hz. But I've seen 2-way bookshelf speakers that measure better than this.

Apologies for my lack of photo hosting, currently, but here is a link to a graph of the JBL M2 Frequency Response for comparison. I've only heard these once, and they were not set up beyond being connected to a CD player on the long 20'+ wall of a closed room. Other than bass booms near boundaries of the room, they sounded the same everywhere, it was truly incredible.

But despite that, a fair argument against the M2 is that compression drivers are not known to convey the 'emotion' (whatever that metric is in music the way a RAAL ribbon can, for example. That's highly subjective, and measurements can't tell us that part. But they do show a speaker that remains completely neutral across the frequency range.

I'm surprised you like them so much on-axis, given how Toole says that the off-axis sound needs to be consistent with the on-axis sound for good sound overall, even when sitting on-axis.

True, the combination of good on and off axis performance yields higher preference ratings in double blind, monophonic, testing.

My point of mentioning the on axis sound as a positive, does not detract from my previous complaint that the sound changed as I moved across the couch, despite the stereo listening situation. Not even the masking effect of stereo could hide that issue.

The D3s do not have a Kevlar midrange driver. It is a carbon fiber weave. It sounds significantly different to the old Kevlar driver.

In addition the new D3 does sound very similar right across the room, the old Kevlar driver did not.

What I found was a slight dip at 5 KHz, and not near as high a peak at 10 HHz. My measurements agree with the sonic footprint of the speaker very well.

The Ds were forward, the D3s are retiring and a very different speaker.

After a listening session at Eagan to my rig there, Phil realized it was time for a change. He has always loved those speakers. He is delighted with his 800 D3 and they sound much more like my designs.

That's interesting that they changed the diaphragm on the mid range. The measurements posted by ADTG don't show much change in frequency response. But obviously these measurements are insufficient to understand how it sounds off axis, so I will take your word for it!

More recent listening has started making me wonder about your point on TL's music renditions. It would seem that the bass propagation of a TL is more closely related to musical instruments than other designs. We've been listening to a lot of live music lately, and have just been in love!

 

[TLS Guy]

According to the graphs above by ADTG, not a single model can maintain even a +/-3dB spec above 200 hz. I agree with you that below 200 hz shows the only significant changes, as all 3 models measure almost identically above 200 hz. But I've seen 2-way bookshelf speakers that measure better than this.

Apologies for my lack of photo hosting, currently, but here is a link to a graph of the JBL M2 Frequency Response for comparison. I've only heard these once, and they were not set up beyond being connected to a CD player on the long 20'+ wall of a closed room. Other than bass booms near boundaries of the room, they sounded the same everywhere, it was truly incredible.

But despite that, a fair argument against the M2 is that compression drivers are not known to convey the 'emotion' (whatever that metric is in music the way a RAAL ribbon can, for example. That's highly subjective, and measurements can't tell us that part. But they do show a speaker that remains completely neutral across the frequency range.

True, the combination of good on and off axis performance yields higher preference ratings in double blind, monophonic, testing.

My point of mentioning the on axis sound as a positive, does not detract from my previous complaint that the sound changed as I moved across the couch, despite the stereo listening situation. Not even the masking effect of stereo could hide that issue.

That's interesting that they changed the diaphragm on the mid range. The measurements posted by ADTG don't show much change in frequency response. But obviously these measurements are insufficient to understand how it sounds off axis, so I will take your word for it!

More recent listening has started making me wonder about your point on TL's music renditions. It would seem that the bass propagation of a TL is more closely related to musical instruments than other designs. We've been listening to a lot of live music lately, and have just been in love!

I can't stress enough that the 800 D3 is a very different sounding speaker to the 800 D and the other iterations of the 800 series which to me sounded pretty much the same.

I have a lot of confidence in my measurements. I listened extensively first and the curves I obtained were exactly what I expected pretty much.

Now as I have often stated it is much easier to get a good frequency response from a 2 way than a three way. That is why my speakers are a three way sleight of hand. Everybody wants three ways, but they should be avoided where possible. When you look at waterfall plots you can always see the discontinuity with every crossover point. Too much reliance on FR alone is not good guidance.

As far as the new FD3 are concerned the off axis response is actually very good, apart from the 90 degree suck out.

As far as FR is concerned there is a slight BBC 'smiley" but moved up not quite an octave. This is why the speaker is a little retiring and not at all forward. If you look at the combined on and off axis HF response the room response will be pretty much spot on, and this correlates with an excellent waterfall response.

The bass is deeper and of better quality in the D3. On the others I could always here port kick in. The bass is much better damped.

As far as my TLs are concerned the bass is more natural. As for the rest the sound is very similar, although I find mine do sound more relaxed without loosing presence.

One last point the lower crossover has been moved up to 500 Hz, which I think helps the load. The Revels, have too lower crossover for the bass drivers and it shows, when you hear it.

 

[Irvrobinson]

One last point the lower crossover has been moved up to 500 Hz, which I think helps the load. The Revels, have too lower crossover for the bass drivers and it shows, when you hear it.

Just out of curiosity, what would I listen for on the Revels? I have a home-made recording of our Steinway upright I use to help get placement right in my system, where I'm playing every note starting with A4 and moving downward. I must not be trained properly, because I can't hear any discontinuity or difference, and I can't hear anything anomalous even when I go to one octave below middle C, which should be right in the range you're talking about. Or I do need test tones to detect it?

 

[Beave]

According to the graphs above by ADTG, not a single model can maintain even a +/-3dB spec above 200 hz. I agree with you that below 200 hz shows the only significant changes, as all 3 models measure almost identically above 200 hz. But I've seen 2-way bookshelf speakers that measure better than this.

So have I. But are the impedance curves really the way to determine poorly matched crossovers and drivers?

Apologies for my lack of photo hosting, currently, but here is a link to a graph of the JBL M2 Frequency Response for comparison. I've only heard these once, and they were not set up beyond being connected to a CD player on the long 20'+ wall of a closed room. Other than bass booms near boundaries of the room, they sounded the same everywhere, it was truly incredible.

Not really a fair comparision, given that the M2 uses active crossovers with programmed DSP filters to achieve that flat response.

 

[TheWarrior]

I can't stress enough that the 800 D3 is a very different sounding speaker to the 800 D and the other iterations of the 800 series which to me sounded pretty much the same.

I have a lot of confidence in my measurements. I listened extensively first and the curves I obtained were exactly what I expected pretty much.

Now as I have often stated it is much easier to get a good frequency response from a 2 way than a three way. That is why my speakers are a three way sleight of hand. Everybody wants three ways, but they should be avoided where possible. When you look at waterfall plots you can always see the discontinuity with every crossover point. Too much reliance on FR alone is not good guidance.

As far as the new FD3 are concerned the off axis response is actually very good, apart from the 90 degree suck out.

As far as FR is concerned there is a slight BBC 'smiley" but moved up not quite an octave. This is why the speaker is a little retiring and not at all forward. If you look at the combined on and off axis HF response the room response will be pretty much spot on, and this correlates with an excellent waterfall response.

The bass is deeper and of better quality in the D3. On the others I could always here port kick in. The bass is much better damped.

As far as my TLs are concerned the bass is more natural. As for the rest the sound is very similar, although I find mine do sound more relaxed without loosing presence.

One last point the lower crossover has been moved up to 500 Hz, which I think helps the load. The Revels, have too lower crossover for the bass drivers and it shows, when you hear it.

Re: 'the 90deg suck out' What do you think is causing this? I site the 'baffle less' tweeter (and that looks to include the midrange as well) as the fault in it's directivity response. I don't think it would be wrong to say that would also cause a loss of amplitude at the crossover frequency for each driver, above the woofers.

I am confused at your issue with 3-way crossovers, which the 800 series is. The Salon2 is a 4 way. But I would like to know as Irv asked, what did you hear? I haven't heard them yet, but hope to change that it the not too distant future...

So have I. But are the impedance curves really the way to determine poorly matched crossovers and drivers?
Not really a fair comparision, given that the M2 uses active crossovers with programmed DSP filters to achieve that flat response.

Using the impedance curves is just one way to analyze that. That is the term for opposition to current when voltage is applied. So if each driver, above the woofer, requires that much change to the current when energized, that could result from a number of issues between components/type of crossover, to the parameters of the drivers themselves. Nonetheless, it is not indicative of good design. I keep mentioning the style of the baffle less drivers for a reason - function following form.

Re: M2 - I felt it was a fair comparison because the Revel's are actually less expensive. They are an active speaker, yes, but Floyd's book and Youtube CIRMMT video show clips of how that speaker measured the same in a home, as well as multiple theaters up to 500 seats - DSP is not that powerful. I gripe about the +/- 3dB spec and used the M2 really to bring that point home. When asking $20k+ for a pair of speakers, I'd want +/- 1dB to be the tolerance, if accurate reproduction at all seats is the goal.

 

[Beave]

Using the impedance curves is just one way to analyze that. That is the term for opposition to current when voltage is applied. So if each driver, above the woofer, requires that much change to the current when energized, that could result from a number of issues between components/type of crossover, to the parameters of the drivers themselves. Nonetheless, it is not indicative of good design. I keep mentioning the style of the baffle less drivers for a reason - function following form.

I still don't see how an impedance curve with large swings in amplitude indicates poor driver/crossover matching or design.

Re: M2 - I felt it was a fair comparison because the Revel's are actually less expensive. They are an active speaker, yes, but Floyd's book and Youtube CIRMMT video show clips of how that speaker measured the same in a home, as well as multiple theaters up to 500 seats - DSP is not that powerful. I gripe about the +/- 3dB spec and used the M2 really to bring that point home. When asking $20k+ for a pair of speakers, I'd want +/- 1dB to be the tolerance, if accurate reproduction at all seats is the goal.

Do you mean JBL, not Revel?

 

[Dennis Murphy]

I still don't see how an impedance curve with large swings in amplitude indicates poor driver/crossover matching or design.

I'm also having trouble following that line of reasoning.

 

[TheWarrior]

I still don't see how an impedance curve with large swings in amplitude indicates poor driver/crossover matching or design.



Do you mean JBL, not Revel?

Sorry, I can't explain it better. But yes, I meant JBL, lol!

 

[Art Vandelay]

I have only high regard for the new B&W line-up, and that now includes the new 700 series. I listened at depth recently to the new 705S2 and was really blown away. This is one of the best 2-way stand mounts I've ever heard.

Comments regarding measured treble response are all misleading too, because the response at 20 degrees off axis is very flat, and not surprisingly this is the intended listening axis. This also applies to all D1, D2, and D3 models.

And yes, the new continuum FST mid is better than the outgoing kevlar, but that alone is not the reason for the whole improvement. B&W engineers have also revised (refined) the balance and voicing by tuning the crossovers and enclosures, and the result is a more seamless and coherent speaker. Not quite as perfect perhaps as the new Wilson Alexia 2, but then nothing quite is. That said, the 800D3 is possibly more environment friendly, and if soundstage is your thing, not much comes close.

Fwiw, I've heard Revel Salon 2's many times and hold them in high regard too. Looking forward to hearing a Salon 3, if and when that transpires.

 

[highfigh]

I still don't see how an impedance curve with large swings in amplitude indicates poor driver/crossover matching or design.

Impedance isn't just a matter of resistance at a particular frequency, it has other components and it causes an amplifier to respond differently, depending on the magnitude of that component. There's Capacitive Reactance, where the impedance decreases as the frequency increases and there's Inductive Reactance, where the impedance increases as the frequency increases. If the drivers/crossover components that are chosen cause the impedance to drop to a level that causes the amplifier to perform poorly, go into protection or suffer damage, it's a poor crossover/speaker design.

 

[AcuDefTechGuy]

I don't think anyone should debate whether any of these speakers sound good or if a D3 sound very different than a D2 or D1 because it's purely subjective.

It's an opinion and everyone has one. And everyone feels confident about their hearing, unless they have hearing impairment.

A speaker could have poor measurements and still sound subjectively great.

3 speakers could measure almost identical and sound subjectively different to some people.

3 speakers could measure every differently and sound subjectively the same to some people.

But if we are just sticking to the FACTS and not opinions, none of these B&W (D1, D2, D3) measure very well in terms of on-axis and off-axis FR when compared to many speakers. Just look at the measurements from Stereophile.

That's a fact.

But sure, aside from the fact that these B&W D3, D2, D1 don't have good on-axis and off-axis FR measurements, they could sound awesome.

 

[<eargiant]

Another fact is that not everything that counts can be counted, and not everything that can be counted counts.

The engineers at Steyning are well aware of this.

 

[Beave]

Impedance isn't just a matter of resistance at a particular frequency, it has other components and it causes an amplifier to respond differently, depending on the magnitude of that component. There's Capacitive Reactance, where the impedance decreases as the frequency increases and there's Inductive Reactance, where the impedance increases as the frequency increases. If the drivers/crossover components that are chosen cause the impedance to drop to a level that causes the amplifier to perform poorly, go into protection or suffer damage, it's a poor crossover/speaker design.

If the 'swings' in magnitude result in really low dips, to, say, 2 Ohms, then I might agree with the statement - because, as you point it, it can stress many/most amplifiers.

If the 'swings' in phase result in really large phase angles, say, +72 degrees, then I might agree with the statement as well - because, as you point out, it can stress many/most amplifiers.

But the way I read the claim, it was that large swings in impedance indicate poor driver selection and/or crossover design. What if those swings are magnitude from 5 Ohms to 40 Ohms in the double humps in the bass in a ported design? What if those swings in phase angle are from +38 degrees to -38 degrees? That's a 'swing' of 76 degrees. Those are large 'swings,' but they won't cause any stress to an amp.

Maybe this is all just about what 'swings' really means? I took it to mean changes. Maybe he meant extremes (low for magnitude, high for phase)?

 

[Beave]

Another fact is that not everything that counts can be counted, and not everything that can be counted counts.

The engineers at Steyning are well aware of this.

How do engineers do their designs if they believe that they can't measure relevant properties?

 

[<eargiant]

How do engineers do their designs if they believe that they can't measure relevant properties?

Possibly by not slavishly adhering to the dogma surrounding the properties they can measure.

 

[Art Vandelay]

Another fact is that not everything that counts can be counted, and not everything that can be counted counts.

The engineers at Steyning are well aware of this.

True. They perform a huge number of 3D measurements in their anechoic chamber, so they know better than anyone how the speaker measures at every point in space. And it's the big picture that matters.

 

[Art Vandelay]

worth a read...https://hometheaterhifi.com/reviews/speaker/floor-standing/bw-803-d3-floor-standing-loudspeaker-review/

 

[TLS Guy]

If the 'swings' in magnitude result in really low dips, to, say, 2 Ohms, then I might agree with the statement - because, as you point it, it can stress many/most amplifiers.

If the 'swings' in phase result in really large phase angles, say, +72 degrees, then I might agree with the statement as well - because, as you point out, it can stress many/most amplifiers.

But the way I read the claim, it was that large swings in impedance indicate poor driver selection and/or crossover design. What if those swings are magnitude from 5 Ohms to 40 Ohms in the double humps in the bass in a ported design? What if those swings in phase angle are from +38 degrees to -38 degrees? That's a 'swing' of 76 degrees. Those are large 'swings,' but they won't cause any stress to an amp.

Maybe this is all just about what 'swings' really means? I took it to mean changes. Maybe he meant extremes (low for magnitude, high for phase)?

The real issue is if the impedance the amp sees at any point is below the DC resistance of the drivers, then that crossover is in resonance for certain. That is bad and should not be tolerated.

 

[highfigh]

If the 'swings' in magnitude result in really low dips, to, say, 2 Ohms, then I might agree with the statement - because, as you point it, it can stress many/most amplifiers.

If the 'swings' in phase result in really large phase angles, say, +72 degrees, then I might agree with the statement as well - because, as you point out, it can stress many/most amplifiers.

But the way I read the claim, it was that large swings in impedance indicate poor driver selection and/or crossover design. What if those swings are magnitude from 5 Ohms to 40 Ohms in the double humps in the bass in a ported design? What if those swings in phase angle are from +38 degrees to -38 degrees? That's a 'swing' of 76 degrees. Those are large 'swings,' but they won't cause any stress to an amp.

Maybe this is all just about what 'swings' really means? I took it to mean changes. Maybe he meant extremes (low for magnitude, high for phase)?

WRT 'swing', think of it as a pendulum.

An inductor or capacitor will present a 90 degree shift in phase but if the value is just wrong, a second order crossover's coil that operates with minimal load below the cap's operating range presents a short. 5 Ohm to 40 Ohm jumps don't really cause problems with an amp, just output, maybe. The impedance peak at Fs keeps the driver from oscillating wildly.