Is this approach better than a coaxial speaker ?

[sarumbear]

I found this

Pseudo-coaxial with narrow directivity (and Horbach-Keele filters)

Well done ! Makes me think of this: http://www.diyaudio.com/forums/multi-way/263353-portal-controlled-dispersion-tripole-4.html -- unfortunately that project seems abandoned And also this: http://www.diyaudio.com/forums/multi-way/90804-large-midrange-ob-scott-g-40.html#post1953221

www.diyaudio.com

And here are the expected lopes

You need an adjustable height stand as well.

 

[thewas]

Perlisten is a good example of MTM and the other is wishful thinking.

Nope, on page 21 and 22 of the pdf the directivity plots are from measurements (which match the simulation ones shown earlier in the same document very well), this guy although hobbyist is constantly doing better engineered designs than most of the pros from the Hifi world.

I electronically translated those parts for interested readers:

Radiation behaviour
The horizontal radiation pattern differs only slightly from the simulated pattern. It depends strongly on
the filter functions and the equalisation of the individual branches, so that a lot of time had to be
a lot of time had to be invested in tuning. The targeted coverage angle of 120° (-6 dB) is maintained over a wide frequency range.
over a wide frequency range.

Vertically there are no surprises. The dispersion behaviour corresponds to the simulated
to the greatest possible extent. Significant side lobes are not present.

 

[sarumbear]

Nope, on page 21 and 22 of the pdf the directivity plots are from measurements (which match the simulation ones shown earlier in the same document very well), this guy although hobbyist is constantly doing better engineered designs than most of the pros from the Hifi world.

I electronically translated those parts for interested readers:

Radiation behaviour
The horizontal radiation pattern differs only slightly from the simulated pattern. It depends strongly on
the filter functions and the equalisation of the individual branches, so that a lot of time had to be
a lot of time had to be invested in tuning. The targeted coverage angle of 120° (-6 dB) is maintained over a wide frequency range.
over a wide frequency range.

View attachment 234202
Vertically there are no surprises. The dispersion behaviour corresponds to the simulated
to the greatest possible extent. Significant side lobes are not present.
View attachment 234203

What is that nope for? Can’t you see the lopes? Or, are you saying that the test sample is not “tuned”?

 

[thewas]

What is that nope for? Can’t you see the lopes? Or, are you saying that the test sample is not “tuned”?

Nope is for that you say it is wishful thinking, as its measurements match excellently its simulation and show excellent vertical directivity. Which lobes are you talking about, the light blue ones at -18 dB level? Also what do you mean with test sample and "tuned"?

 

[sarumbear]

Nope is for that you say it is wishful thinking, as its measurements match excellently its simulation and show superb vertical directivity. Which lobes are you talking about, the light blue with -15 dB level?

It’s wishful thinking is that the designer expected a useable speaker with his design. He may have successfully designed and then built something but it is not not good enough for the job.

Knowing wave theory I expected to see those lopes (shades of red) on horizontal response by only looking at the pictures and there they are on the measurements.

 

[thewas]

It’s wishful thinking is that the designer expected a useable speaker with his design. He may have successfully designed and then built something but it is not not good enough for the job.

For which job it is not good enough?

Knowing wave theory I expected to see those lopes (shades of red) from the design I see on the pictures and there they are on the measurements.

Please show exactly which regions on the plots you mean and also measurements of a Hifi loudspeaker with so wide horizontal and narrow directivity which are at the same time so smooth.

 

[sarumbear]

For which job it is not good enough?

Speakers for home use

Please show exactly which regions on the plots you mean and also measurements of a Hifi loudspeaker with so wide horizontal and narrow directivity which are at the same time so smooth.

Marked blue. 6dB change within a few degrees.

You asked smooth polar response?

Official KEF Ci5160REF -THX spinorama

I am in the process of replacing my JBL C222HP LCR speakers with KEF Ci5160REF -THX. I asked KEF to send me the spinorama chart. It took them a while but the delay turned out to due to their updating the model to THX spec. I thought the chart will be valuable information for the forum. I have...

www.audiosciencereview.com

 

[thewas]

Speakers for home use


Marked blue. 6dB change within a few degrees.
View attachment 234209
You asked smooth polar response?

Official KEF Ci5160REF -THX spinorama

I am in the process of replacing my JBL C222HP LCR speakers with KEF Ci5160REF -THX. I asked KEF to send me the spinorama chart. It took them a while but the delay turned out to due to their updating the model to THX spec. I thought the chart will be valuable information for the forum. I have...

www.audiosciencereview.com

First of all the step from light red to dark red is 3 and not 6 dB.
Second your KEF are not a wide horizontal and narrow vertical loudspeaker but symmetrical which is much easier to achieve.
Third their comparative polar plots are not smoother but rather the opposite:

 

[sarumbear]

First of all the step from light red to dark red is 3 and not 6 dB.

I don’t know how you read that plot but I read +3dB and -3dB, which is 6dB.

Second your KEF are not a wide horizontal and narrow vertical loudspeaker but symmetrical which is much easier to achieve.

the title of the post is:

Is this approach better than a coaxial speaker ?​

Third their comparative polar plots are not smoother but rather the opposite:

I also don’t know where you get that plot but it belongs to a different speaker!

 

[thewas]

Speakers for home use

I see it the opposite way, narrow vertical directivity can be a big advantage at homes especially at the typically too large listening distances as it reduces the influence of the floor and ceiling reflections, that's why companies like Perlisten (or Genelec with its S360) go the extra effort for it.

I don’t know how you read that plot but I read +3dB and -3dB, which is 6dB.

I see that dark red colour covers the region from +3 to 0 dB and light red from 0 dB to -3 dB so I remain that their difference is 3 dB.

the title of the post is:

Is this approach better than a coaxial speaker ?​

Your comment that initiated this discussion between us was that this "is wishful thinking" which it isn't because you probably saw the weird driver configuration and distances without reading the engineering behind.

I also don’t know where you get that plot but it belongs to a different speaker!

Because it is the only KEF in-wall coaxial I know for which such detailed polar measurements exist, if you have polars for your larger model please post them.

 

[sarumbear]

I see it the opposite way, narrow vertical directivity can be a big advantage at homes especially at the typically too large listening distances as it reduces the influence of the floor and ceiling reflections.

A CBT design can achieve that using a much simpler design.

But your comment that initiated this discussion was that this "is wishful thinking" which it isn't because you probably saw the weird driver configuration and distances without reading the engineering behind.

I skimmed it, I admit, but I categorically ignore such designs where things have to be “tuned” and/or linearity suffers. They are freak designs that at best tangent to wave theory. Something always gives when you ignore physics. Engineering can never surpass physics.

Because it is from the same series and the only one I know for which detailed measurements exist, if you have polars for your larger model please post them.

SPINORAMA is the same data displayed in a different chart format. You can see that the response is smooth at all angles. That means there are no lopes.

 

[thewas]

A CBT design can achieve that using a much simpler design.

A good implementation of a CBT isn't really simpler, also the horizontal of the JBL you chose is clearly inferior:

I skimmed it, I admit, but I categorically ignore such designs where things have to be “tuned” and/or linearity suffers. They are freak designs that at best tangent to wave theory. Something always gives when you ignore physics. Engineering can never surpass physics.

What "freak design" are you talking about and where exactly do you see the "wave theory abused" by it, this design cleverly uses physics (similar to CBT) and thus has not really "weird" crossover functions which "make linearity suffer" but typical Horbach und Keele (see CBT) and Linkwitz/Riley ones:

SPINORAMA is the same data displayed in a different chart format. You can see that the response is smooth at all angles. That means there are no lopes.

The spinorama you posted shows explicitly only the +-90° degrees, still you see that the polars of the other KEF one I showed are inferior (also because the 10 dB per line scaling makes it look smoother than a 3 dB per colour).

 

[TurtlePaul]

I see it the opposite way, narrow vertical directivity can be a big advantage at homes especially at the typically too large listening distances as it reduces the influence of the floor and ceiling reflections, that's why companies like Perlisten (or Genelec with its S360) go the extra effort for it.

There is a big difference between some of the polar plots posted above and the S360. The S360 control directivity +/- 30 degrees in the vertical. There are some plots above that are +/- 10 degrees. When you are approaching that narrow, then standing up results in all the highs going away. That is much less useful for in-home situations (you need to sit strictly at tweeter level to get anything useable).

 

[OWC]

As long as you cross the speaker below the first inference mode (path length difference between the two speakers), any speaker is the same as a coaxial.
In practice we only have to concern the vertical off-axis response as well as the vertical directivity.

It's extremely debatable how significance that even is.
I have never seen a lot of PROPER research done on this.
In fact, there is plenty of research showing that the vertical sensitivity of human hearing is pretty poor.

Also, since ceiling-floor distance in most rooms is quite low (around 2,6 meters or so), as well as the floor-speaker distance, it's also debatable how much of this nice "perfect" pattern will stay intact. Especially since any system will always introduce other problems and compromises.

It's also only needed for the mid-high frequencies.
For lower frequencies it's a total waste, since the wavelengths are much much bigger, as well as all the room modes that just ruin everything in an instance.
So in the end when it comes down to the lower frequencies, people have created a "solution" that looks cool on paper, but in practice it's not beneficial.

 

[Tangband]

As long as you cross the speaker below the first inference mode (path length difference between the two speakers), any speaker is the same as a coaxial.
In practice we only have to concern the vertical off-axis response as well as the vertical directivity.

If this is true, one could instead of a uniQ driver use a 4 inch driver and a tweeter with minimal faceplate and mount them very close and do the crossover at 2 kHz . In this way , you should avoid the bassdriver modulating the tweeter, as Erin has shown before. And no waveguide needed…
I have two drivers in mind, a 4 inch SB PAC driver and a low distortion monacor tweeter ( see picture )

But the fact is - I feel thats there is something more with the uniQ drivers, maybe its the very low distortion tweeter they use ?

 

[OWC]

If this is true, one could instead of a uniQ driver use a 4 inch driver and a tweeter with minimal faceplate and mount them very close and do the crossover at 2 kHz . In this way , you should avoid the bassdriver modulating the tweeter, as Erin has shown before.

But the fact is - I feel thats something more with the uniQ drivers, maybe its the low distortion tweeter they use ?

That is doppler distortion.
So technically one could also prevent the woofer from have to much excursion

One of the compromises of coaxial speakers, is the shape of the cone.
For rigidity purposes, as well having a nice off-axis response the woofer needs a different shape than what is suitable for a waveguide for the tweeter.
For the tweeter a CD-waveguide or a Oblate Spheroid Waveguide is needed to have proper constant directivity.
Which doesn't really suit very well with what is needed for the woofer in sense of unwanted cone resonances.

As for the wavelength difference, this all depends what exactly you're looking for.
The frequency of this difference can be calculated with;

f = v / (2·D·sin(a) )

Where;
f = frequency
v = speed of sound
D = distance between sound sources in meters (center-to-center distance)
a = angle in degrees.

So for an angle of 90 degrees (= max difference in wavelength), we can skip the sin(a) term, since it will be 1.
A center-to-center distance of 15cm gives us around 1146Hz.
Although in practice the frequency already starts dipping before.
So it's not easy to get this super well, especially when one wants to use a nice waveguide for the tweeter.
Which is a MUST for a good directivity.

I am not concerned about tweeter distortion, these days this distortion is low enough from any decent manufacturer.

 

[test1223]

The Monitor Audio Concept 50 was discussed here with some comments on the unusual driver arrangement. https://www.audiosciencereview.com/forum/index.php?threads/monitor-audio-concept-50-announced.34117/

The main problem of such a design is the tweeter which has to play very low to avoid the discussed lopes and to place everything as close as possible on the baffle. This is a very hard task since the distances have to be very small. If you think of infinite small distances you get a usual coax speaker design with only one midrange driver.

I found this

edit: that imagine is also from the pdf lol

Pseudo-coaxial with narrow directivity (and Horbach-Keele filters)

Well done ! Makes me think of this: http://www.diyaudio.com/forums/multi-way/263353-portal-controlled-dispersion-tripole-4.html -- unfortunately that project seems abandoned And also this: http://www.diyaudio.com/forums/multi-way/90804-large-midrange-ob-scott-g-40.html#post1953221

www.diyaudio.com

And here are the expected lopes

Nope, on page 21 and 22 of the pdf the directivity plots are from measurements (which match the simulation ones shown earlier in the same document very well), this guy although hobbyist is constantly doing better engineered designs than most of the pros from the Hifi world.

The strongest lopes will occur with this speaker if you have a look at frequency responses where the horizontal and vertical angle aren't 0°. The biggest relative distances of the drivers create the worst lopes.

 

[Prana Ferox]

These designs are gimmicks. The buyer thinks "Hmm, more drivers, clearly this is worth spending more money than a speaker with less drivers." That's the gimmick.

 

[Plcamp]

There is still all kinds of movement around the tweeter, you can’t have a smooth waveguide either.

I have had this question bugging me for a while…what if you created a smooth waveguide using material that was semi-acoustically transparent. Screen material can be purchased with varying % opaqueness as one example.

Could midranges drive through such a screen, while the screen still maintains enough directivity control for the tweeter? A kind of co-entrant horn?

 

[okok]

are you the Tangband speaker maker?

driver maker