The quest for my hyper speaker - Very Large room dilemma

[aliqaz]

The Everests are not really constant directivity horns. Some may arrest me for this, but IMO they are designed more for looks than absolutely best sound quality. Not saying the can't sound good, but they have some obvious compromises. One is the mentioned directivity. The Jubilees have a top horn with much more uniform directivity and therefore will give a more even frequency response in the range of that horn when placed in the actual room combined with the fact that it's active, but the weaknesess are the following:

- They loose vertical directivity quite high in frequency
- They use a 2" exit which exhibits higher distortion and a bit more narrowing in the highs
- The bass bin doesn't offer good directivity control and is somewhat uneven on-axis in parts of the region. The latter is something Klipsch may avoid with a very low crossover though, but that again leads to issues in relation to the vertical directivity because the top horn doens't maintain it low enough.

Uniform directivity is essential and should ideally be that down to the Schroeder frequency. This leads to a much less coloration from surfaces and more a correct tonality. This is actually more important that a super even frequency response on-axis anechoic. Especially in rooms with no or little acoustic treatment. Most speakers recommmended here don't fit that bill and will not measure that well placed in a normal room. That being said, in this big room you will avoid many acoustic problems. At the same time, some other will arise but they are easier to deal with.

With a horn speaker and directivity control, we don't get away from large dimensions.

A very interesting and education analysis. Do you have any recommendations?

 

[fpitas]

The Everests are not really constant directivity horns. Some may arrest me for this, but IMO they are designed more for looks than absolutely best sound quality. Not saying the can't sound good, but they have some obvious compromises. One is the mentioned directivity. The Jubilees have a top horn with much more uniform directivity and therefore will give a more even frequency response in the range of that horn when placed in the actual room combined with the fact that it's active, but the weaknesess are the following:

- They loose vertical directivity quite high in frequency
- They use a 2" exit which exhibits higher distortion and a bit more narrowing in the highs
- The bass bin doesn't offer good directivity control and is somewhat uneven on-axis in parts of the region. The latter is something Klipsch may avoid with a very low crossover though, but that again leads to issues in relation to the vertical directivity because the top horn doens't maintain it low enough.

Uniform directivity is essential and should ideally be that down to the Schroeder frequency. This leads to a much less coloration from surfaces and more a correct tonality. This is actually more important that a super even frequency response on-axis anechoic. Especially in rooms with no or little acoustic treatment. Most speakers recommmended here don't fit that bill and will not measure that well placed in a normal room. That being said, in this big room you will avoid many acoustic problems. At the same time, some other will arise but they are easier to deal with.

With a horn speaker and directivity control, we don't get away from large dimensions.

Can you share the measurements of the DD67000?

 

[GXAlan]

@aliqaz
Respecting your privacy, we won’t be able to recommend high end retailers are near you.

That said, I am curious to your average SPL’s in your other JBL 4349 room or your experience with the Kanta’s in the giant room.

Maybe all of our thoughts about going for maximum dynamics and SPL are unnecessary and it turns out that even the 4349 will hit your desired SPLs in that room.

 

[Bjorn]

A very interesting and education analysis. Do you have any recommendations?

I'm afraid I'm higly biased since I develop speakers.
Besides a uniform directivity I focus on areas like limiting the vertical dispersion, avoiding the floor bounce, getting the crossover far away from the most sensitive area, active crossover with amp to each driver, low distortion, and dynamics.

I've found that commercial speakers are generally only good in some of theses areas but not all.

 

[GXAlan]

The Everests are not really constant directivity horns. Some may arrest me for this, but IMO they are designed more for looks than absolutely best sound quality.

JBL kept the “DD” terminology of Defined Directivity.

The original being the Paragon and the second being the original Everest. Neither would meet today’s standards for audio, but as an owner of the JBL S/2600, I can say that those asymmetrical horns work great. It’s a big compromise in quality but a good compromise in some environments. I would compare it to a Revel standard by saying that Revel gives you an A+ in the main listening position, but if you sit right in front of one of the speakers, it’s a D-. The dispersion on the S/2600 is completely inconsistent as you move off axis, but I still would say that main listening position is a B+ and it drops to a B- at the incorrect seating positions.

I think the closest way to justify this in science is that we know that stereo listening diminishes the discrimination between good and bad speakers from Dr. Toole’s work. With something like a Revel, once you move so far from center that it is essentially mono, the DD JBLs which can preserve stereo at more extreme positions catches up.

Besides the primary horn not being a constant directivity as you mention, the two woofers aren’t identical either. One 15” is crossed over at 700Hz and the other is 150 Hz. Looking at the marketing, JBL only states that the UHF driver is a constant directivity, so your insights on the main compression driver not meeting that standard is true.

 

[Bjorn]

Can you share the measurements of the DD67000?

I have only seen on-axis measurement and a waterfall of the DD67000. But there's much we can know based on physics and knowledge about speaker design.

Here's the on-axis measurement measured by Hifi News:

The uneven response in the highs is as expected. When you have two vertical drivers with this much spacing and a crossover at 20 kHz, the result is polar lobing and comb filtering. That's unavoidable.

The horn is very short in regards to directivity control. Result is that it will only provide a constant vertical directivity in a small frequency area. Horizontally I don't how well the horn measures off-axis. It would be interesting to see. Crossing from a horn over to a front firing at 850 Hz isn't trivial. Front firing woofers are omni in the lows and than gradually narrows at higher frequencies.
30 Hz:

331 Hz:

Above approximately 400- 500 Hz, the woofers will exhibit polar lobing and which get's worse the higher you go up.
501 Hz:

840 Hz:

But I think the main problem is the lack of a uniform directivity from front firing woofers.
I haven't experienced much with dual 15" woofers next to each other, but I've worked quite a bit with crossing a large horn to a single 15" front firing woofer and which should be similar considering the 2.5 design. I've used a steep crossover at 600 Hz which makes it easier. But it's challenging because you are going from a horn with constant directivity to a woofer that isn't constant. The woofer is therefore much more colored in the response from the room. Crossing over to a midbass horn with similar directivity is far more seamless and yields a much more even response. The drawback here is size.

BTW. We can see how the horizontal polar of 15" will be similar to, by looking at the vertical polar of the JBL 4638.

Whilst such off-axis measurements are limited compared to a polar shown as sonocrame, we can still cleary see some issues here. Edit: This will be different with a crossover to a single 15" in a 2.5 design.

When the speaker is passive with no signal aligment, that means the sound from the different drivers are not reaching the ear at the same time. Personally I find that's an highly audible drawback.

The impulse response from the previous DD66000 model is here:

As a side note. It's sad how difficult it is these days to get proper measurements from the speaker brands. I highly doubt JBL would be willing to share proper polars. Quite the contrast to the past when i.e. Don Keele worked there and showed everything. He drew everything by hand by the way.

 

[Galliardist]

I have little to add here, except two show experiences in big rooms, probably not suitable:

Kyron Audio - listened to something of theirs on a massive space, but have no idea what they were actually running.

And seeing Tannoys mentioned here reminded me of the Kingdom speaker which I’ve also heard fill a room that the Westminsters didn’t really cope with.

Both options very expensive…

 

[fpitas]

I have only seen on-axis measurement and a waterfall of the DD67000. But there's much we can know based on physics and knowledge about speaker design.

Here's the on-axis measurement measured by Hifi News:
View attachment 282780

The uneven response in the highs is as expected. When you have two vertical drivers with this much spacing and a crossover at 20 kHz, the result is polar lobing and comb filtering. That's unavoidable.

The horn is very short in regards to directivity control. Result is that it will only provide a constant vertical directivity in a small frequency area. Horizontally I don't how well the horn measures off-axis. It would be interesting to see. Crossing from a horn over to a front firing at 850 Hz isn't trivial. Front firing woofers are omni in the lows and than gradually narrows at higher frequencies.
30 Hz:
View attachment 282782

331 Hz:

View attachment 282783

Above approximately 400- 500 Hz, the woofers will exhibit polar lobing and which get's worse the higher you go up.
501 Hz:
View attachment 282784

840 Hz:
View attachment 282781


But I think the main problem is the lack of a uniform directivity from front firing woofers.
I haven't experienced much with dual 15" woofers next to each other, but I've worked quite a bit with crossing a large horn to a single 15" front firing woofer and which should be similar considering the 2.5 design. I've used a steep crossover at 600 Hz which makes it easier. But it's challenging because you are going from a horn with constant directivity to a woofer that isn't constant. The woofer is therefore much more colored in the response from the room. Crossing over to a midbass horn with similar directivity is far more seamless and yields a much more even response. The drawback here is size.

BTW. We can see how the horizontal polar of 15" will be similar to, by looking at the vertical polar of the JBL 4638.
View attachment 282786View attachment 282785

Whilst such off-axis measurements are limited compared to a polar shown as sonocrame, we can still cleary see some issues here. Edit: This will be different with a crossover to a single 15".

When the speaker is passive with no signal aligment, that means the sound from the different drivers are not reaching the ear at the same time. Personally I find that's a highly audible drawback.

The impulse response from the previous DD66000 model is here:
View attachment 282787

As a side note. It's sad how difficult it is these days to get proper measurements from the speaker brands. I highly doubt JBL would be willing to share proper polars. Quite the contrast to the past when i.e. Don Keele worked there and showed everything. He drew everything by hand by the way.

The second woofer crosses at 150Hz to compensate baffle step. I wonder if it really makes much difference up at the horn crossover.

 

[Bjorn]

The second woofer crosses at 150Hz to compensate baffle step. I wonder if it really makes much difference up at the horn crossover.

There's less lobing when they are used that way. Above 200 Hz, two 15" woofers next to each start to exhibit polar lobing. But polar lobing at low frequencies is completely different vs at a sensitive area. At 850 Hz the issue increases.

Either way, there's no uniform directivity from such front firing woofer. Neither vertically or horizontally. Using a hornloaded midbass is a far better design in every way. It lowers the distortion as well.

 

[hex168]

Thank you for this additional information. Imo you do not need any "room treatment"; you just need good quality, very-high-SPL-capable speakers whose reflections have the same spectral balance as their direct sound. And imo this is why you want to minimize any spectral discrepancy between the two:

"If the spectra of the direct and reflected sounds are significantly different, the reflections are likely to be more noticeable, from subtle timbral effects up to a premature breakdown of the precedence effect." - Floyd Toole



Some years ago I had a home demo room that was 33 feet (10 meters) long. I'm a dealer for a line-source-approximating speaker (SoundLab faceted-curved fullrange electrostats), so I took pink noise measurements of a point-source-approximating loudspeaker and the SoundLabs at 1 meter and also back at 8 meters (the practical maximum with the SoundLabs set up correctly). From 1 meter to 8 meters, the point-source-approximating speaker's SPL fell off by 11 decibels. Point source theory predicts a 6 decibel falloff for each doubling of distance, not counting the contribution of in-room reflections. That would theoretically have been an 18 decibel falloff (three doublings of distance), so evidently the in-room reflections were making a 7 decibel contribution back at 8 meters.

Line source theory predicts a 3 dB falloff per doubling of distance. So that would be 9 decibels of falloff, not counting the in-room reflections. The line-source-approximating dipolar SoundLabs fell off by only 4 decibels over the same distance, indicating that the in-room reflections were contributing about 5 decibels at 8 meters.

So in my experience, in a suitable room, a good line-source-approximating loudspeaker's output really does fall off more slowly with distance.

Unfortunately I do not think you can expect a line-source-approximating loudspeaker to behave as well in your room because of your tall ceilings. Let me explain:

The floor and ceiling reflections extend the virtual height of a line-source-approximating loudspeaker, effectively trippling the array height according to Sound Lab designer Roger West. But in order for this to happen, the bottom of the line-source-approximating speaker needs to be very close to the floor, and the top needs to be very close to the ceiling. Your ceiling is so tall that, unless you have custom hyper-tall line-source-approximating speakers made, the ceiling will be too far away for its reflection to effectively extend the array in that direction.

Also, most "line array" type loudspeakers have very wide horizontal dispersion and therefore significant sidewall interaction, and in a room the size of yours I think the resulting abundance of early-onset reflections (combined with the direct-to-reverberant ration being reduced by the listening distance anyway) would work against clarity.

So in general I would not recommend a line array loudspeaker system for your room.

"So in general I would not recommend a line array loudspeaker system for your room." Unless it is very tall. The link below will put some numbers on it:

Design Guidelines for Practical Near Field Line Arrays ...​

 

[Mr. Widget]

Using a hornloaded midbass is a far better design in every way. It lowers the distortion as well.

OK, but at 10 cubic feet the DD67000 is too large for most applications already. These things are meant to go in people's homes... even the OP with his unusually large room would probably balk at a pair of speakers large enough to house a midbass horn.

 

[Bjorn]

The original being the Paragon and the second being the original Everest. Neither would meet today’s standards for audio, but as an owner of the JBL S/2600, I can say that those asymmetrical horns work great. It’s a big compromise in quality but a good compromise in some environments. I would compare it to a Revel standard by saying that Revel gives you an A+ in the main listening position, but if you sit right in front of one of the speakers, it’s a D-. The dispersion on the S/2600 is completely inconsistent as you move off axis, but I still would say that main listening position is a B+ and it drops to a B- at the incorrect seating positions.

I think the closest way to justify this in science is that we know that stereo listening diminishes the discrimination between good and bad speakers from Dr. Toole’s work. With something like a Revel, once you move so far from center that it is essentially mono, the DD JBLs which can preserve stereo at more extreme positions catches up.

Besides the primary horn not being a constant directivity as you mention, the two woofers aren’t identical either. One 15” is crossed over at 700Hz and the other is 150 Hz. Looking at the marketing, JBL only states that the UHF driver is a constant directivity, so your insights on the main compression driver not meeting that standard is true.

I hope you aware that Revel has collapsing horizontal directivity (not constant), vertical lobing and superposition between drivers, floor bounce, no time alignment between drivers, cabinet diffraction issues, and higher modulation distortion compared to what achievable. If such a speaker is A+, I wonder that grade we should give better designs!

 

[Keith_W]

Kyron Audio - listened to something of theirs on a massive space, but have no idea what they were actually running.

They are running DEQX with several Class D amps that they constructed themselves. The reason you don't actually see a DEQX is because it has been stripped of its chassis and only its internals are being used.

 

[Galliardist]

They are running DEQX with several Class D amps that they constructed themselves. The reason you don't actually see a DEQX is because it has been stripped of its chassis and only its internals are being used.

I remembered they use DEQX but not much more.

IIRC they also custom install their main product, hence huge prices.

DEQX are revising their product range at the moment which may also affect Kyron.

 

[GXAlan]

I hope you aware that Revel has collapsing horizontal directivity (not constant), vertical lobing and superposition between drivers, floor bounce, no time alignment between drivers, cabinet diffraction issues, and higher modulation distortion compared to what achievable. If such a speaker is A+, I wonder that grade we should give better designs!

The context was just relative to the JBL’s and decision between great sound at MLP or very good sound broadly. If you give the Revel a B+ instead, you’d just have to drop the Everests down too.

Polar plot and on axis frequency response/distortion data.

https://support.jbl.com/on/demandware.static/-/Sites-masterCatalog_Harman/default/dw99ad9f67/pdfs/S2600,S3100%5EJBL%5Eom%5E%5E%5E%5E.pdf

Should sound horrible, but it doesn’t!

 

[fpitas]

As a registered heretic, I don't particularly "believe" in constant directivity. Lots of great speakers have rising directivity with frequency. Kind of inevitable if you use a 1" dome tweeter, yet no one carps.

 

[sarumbear]

Audioholics recently gave a good review to these: https://www.audioholics.com/tower-speaker-reviews/rtj-410-rtj18sub

the picture is confusing at first as there’s another set of speakers in the room. The RTJ-410 units are sandwiched between the rtj18 subs above and below them.

they measured an in-room F3 below 8Hz, and over 120dB of output capability

You need 25cm peak to peak cone movement to reach those figures. Can you imagine those woofers managing that?

 

[RayDunzl]

Nobody mentioned the Beolab 90.

Beolab 90 - Ultimate High-End 8,200W Speakers | B&O

Perhaps the world's most advanced digital loudspeaker. Bang & Olufsen's flagship stereo speakers adapting seamlessly to your living space.

www.bang-olufsen.com

Frequency Range​

< 12 to > 43,000 Hz

Maximum Sound Pressure Level @1m​

126 dB SPL

Bass Capability​

123 dB SPL

Advanced Sound Features​

Active Room Compensation Adaptive Bass Linearization Beam Width Control
Beam Direction Control (5 sides) Thermal Protection

Power Amplifiers – Custom designed for Beolab 90
7 x Bang & Olufsen ICEpower 300 watts (tweeter)
7 x Bang & Olufsen ICEpower 300 watts (midranges)
4 x Heliox AM1000-1 1000 watts (woofers)

Recommended Room Size
30-200 m² / 300-2000 ft²

Bang & Olufsen BeoLab 90 loudspeaker Measurements

Sidebar 3: Measurements The BeoLab 90 is so massive that, rather than having it shipped to my home, I traveled to Kal Rubinson's place to take some measurements. I used SMUGSoftware's Fuzzmeasure 3 app running on my MacBook Pro, along with an Earthworks QTC-40 microphone and a FireWire-connected...

www.stereophile.com

 

[Ciobi69]

Nobody mentioned the Beolab 90.

Beolab 90 - Ultimate High-End 8,200W Speakers | B&O

Perhaps the world's most advanced digital loudspeaker. Bang & Olufsen's flagship stereo speakers adapting seamlessly to your living space.

www.bang-olufsen.com

Frequency Range​

< 12 to > 43,000 Hz

Maximum Sound Pressure Level @1m​

126 dB SPL

Bass Capability​

123 dB SPL

Advanced Sound Features​

Active Room Compensation Adaptive Bass Linearization Beam Width Control
Beam Direction Control (5 sides) Thermal Protection

Power Amplifiers – Custom designed for Beolab 90
7 x Bang & Olufsen ICEpower 300 watts (tweeter)
7 x Bang & Olufsen ICEpower 300 watts (midranges)
4 x Heliox AM1000-1 1000 watts (woofers)

I think they are very expensive

 

[RayDunzl]

I think they are very expensive

I couldn't identify anything "cheap" in his room picture.