Visualizing How Different Loudspeaker LF Directivity Patterns Couple to Room Modes

[thewas]

If it is correct that D&D uses cardioid all the way down and Kii Three stops at 80 Hz, it's interesting to compare them.

No, the D&D 8c is only cardioid up to approximately 100 Hz so the difference you show from EAC LP responses is probably more due to different placement.

 

[tw99]

No, the D&D 8c is only cardioid up to approximately 100 Hz so the difference you show from EAC LP responses is probably more due to different placement.

Also, that D&D graph shows that most likely, the front wall distance was set incorrectly in the control app.

 

[MKR]

Not to mention the 8C has no EQ, so very impressive, and I think plot of Bjorn speaker is with EQ. Also 8C is not full range cardioid … wasn’t that the point of this discussion?

Bjorn also conspicuously left out the quote from Erin that went along with the graph:
“the thing I notice here, though, is just how well behaved the room modes are above 100Hz. There are a few but they are much less severe than other in-room measurements I have taken of past speakers. One has to think this is attributed to the cardioid nature of the 8c speakers.”

And some interesting papers on the topic:
https://secure.aes.org/forum/pubs/conventions/?elib=12401
https://aes2.org/publications/elibrary-page/?id=13120
https://aes2.org/publications/elibrary-page/?id=17270

 

[Bjorn]

There are many papers that claims otherwise to this your statement. For example, here are some the decay time graphs of room modes ("low frequency reverberation") for monopoles and cardioids (from paper by Ferekidis and Kempe, Controlling the Mode Excitation of Rooms by using Multiple Low Frequency Cardioids in Multichannel Systems). From this investigation follows that the cardioid provides faster modes decay and more even room modes excitation

View attachment 425295

====================
View attachment 425296
=======================

View attachment 425297

The measurements are conducted in a reverberation chamber! So it actually backs up what I've been saying all the time: In a large enough venue directivity control works and that's where cardioid haven been tradionally used and where it can make sense. But in a small room the directivity is swamped by room modes.

 

[Bjorn]

Not to mention the 8C has no EQ, so very impressive, and I think plot of Bjorn speaker is with EQ. Also 8C is not full range cardioid … wasn’t that the point of this discussion?

Bjorn also conspicuously left out the quote from Erin that went along with the graph:
“the thing I notice here, though, is just how well behaved the room modes are above 100Hz. There are a few but they are much less severe than other in-room measurements I have taken of past speakers. One has to think this is attributed to the cardioid nature of the 8c speakers.”

And some interesting papers on the topic:
https://secure.aes.org/forum/pubs/conventions/?elib=12401
https://aes2.org/publications/elibrary-page/?id=13120
https://aes2.org/publications/elibrary-page/?id=17270

First of all, I apologize for indicating D&D 8C were running cardioid lower in frequency. I should have checked that.

But response is impressive?! No, it's absolutely not. That's a poor response in the area of 100 Hz t0 1000 Hz where the cardioid definetly works. It doesn't matter what Erin writes when the graph show otherwise. Perhaps he has never tested any real good speaker designs and lacks references. I don't know, but that's certainly not a great response or "well behaved room modes above 100 Hz". There's a broad cancellation between 150 Hz and 500 Hz. So not good at all.

The best way to achieve an even response for frequencies below 100 Hz is using subwoofer(s), acoustic treatment or a combination. Both have their advantages and disadvantages.

 

[AudioJester]

100Hz to Schroeder is so difficult. Subs wont help here. Kii three and W371 seems to tame it best. But its so hit and miss, as in may not work in your room.

 

[Bjorn]

100Hz to Schroeder is so difficult. Subs wont help here. Kii three and W371 seems to tame it best. But its so hit and miss, as in may not work in your room.

Not necessarily. But it requires a really good speaker design and some work with placement.

No cardioid or EQ here.

 

[Dmitrij_S]

The measurements are conducted in a reverberation chamber! So it actually backs up what I've been saying all the time: In a large enough venue directivity control works and that's where cardioid haven been tradionally used and where it can make sense. But in a small room the directivity is swamped by room modes.

The size of the reverberation chamber corresponds to the size of a standard living room (the frequency of the first mode is ~37-38 Hz, i.e. the largest room size is ~4.8 m). Therefore, the conclusions of the article are also applicable to standard living room (actually, the situation in the reverberation chamber is worse than in a living room).

 

[AudioJester]

Not necessarily. But it requires a really good speaker design and some work with placement.

No cardioid or EQ here.
View attachment 425444

What speaker is this? Looks fantastic! Iam guessing room is heavily treated?

 

[MKR]

Not necessarily. But it requires a really good speaker design and some work with placement.

No cardioid or EQ here.
View attachment 425444

Can you please show this on a 50db vertical scale?

 

[MKR]

First of all, I apologize for indicating D&D 8C were running cardioid lower in frequency. I should have checked that.

But response is impressive?! No, it's absolutely not. That's a poor response in the area of 100 Hz t0 1000 Hz where the cardioid definetly works. It doesn't matter what Erin writes when the graph show otherwise. Perhaps he has never tested any real good speaker designs and lacks references. I don't know, but that's certainly not a great response or "well behaved room modes above 100 Hz". There's a broad cancellation between 150 Hz and 500 Hz. So not good at all.

The best way to achieve an even response for frequencies below 100 Hz is using subwoofer(s), acoustic treatment or a combination. Both have their advantages and disadvantages.

Erin has never tested any real good speaker designs and lacks references? Are you serious?

 

[thewas]

Not necessarily. But it requires a really good speaker design and some work with placement.

No cardioid or EQ here.

Name price and size and do comparative measurements to such a cardioid in the same room and location or send it to Erin, otherwise its just apple vs pears marketing.

 

[Bjorn]

The size of the reverberation chamber corresponds to the size of a standard living room (the frequency of the first mode is ~37-38 Hz, i.e. the largest room size is ~4.8 m). Therefore, the conclusions of the article are also applicable to standard living room (actually, the situation in the reverberation chamber is worse than in a living room).

It's reverberation chamber....Not applicable to a standard living room by any means.
We need to see comparison measurement in a normal room where both have been optimized for best placement. That was done in one paper and I've already shared the result.

It's been said here that cardioid bass has a big advantage in regards to minimizing standing waves vs a monopole. And now it seems we are down to small differences in decay in a reverberation chamber. Just looking at decay or resonances is challenging for several reasons:
1. You need to make sure the monopole is really well built. Especially many commercial bass reflex subwoofers would do poorly here because of typical too small enclosures and or too short ports.
2. The monopole will go lower in frequency unless they are equalized similar. And lower bass means triggering lower room modes and you harmonics on top. That's the reason why some see an improved response with dipole speakers. The speaker simpy doesn't trigger the lowest room modes because of the dipole cancellation. Less deep=less bass problems
3. If by any chance the cardioid would have a slight shorter decay when everything is equal, this needs to weight against the downside of less SPL and higher distortion.

I'm planning to some comparison in an actual room with cardioid sub vs monopole and where the subwoofers are identical. Not sure when I'll get to it, but it's on the do list.

 

[Bjorn]

What speaker is this? Looks fantastic! Iam guessing room is heavily treated?

It's the Vera Audio Coherence 12 prototype. Hardly any treatment at all and actually a difficult placement/room (pic below). No sub was used of course.


There's an introduction thread here FIY: https://www.audiosciencereview.com/...a-high-quality-speaker-many-can-afford.58054/

Sure, I was lucky where the lowest room modes or I move out of it much as possible. So it's not going to measure like that in every room or with every placement. But above approximately 120 Hz, the speaker will generally yield a much more even response than a cardioid satelite speaker. Actually, all of my speaker designs will do that. It comes down to the design of the speaker.

Here's a measurement of a horn speaker indoor. Crossed at sub at 90 Hz, and of course no room EQ.
Left and right channel.

Or a measurement of a CBT speaker indoor where one placement in up against the front wall and the other with a good distance from it:

If one struggles with a uneven response below 100 Hz and can't do physical treatment, using separate subwoofer(s) is by far the best option. The drawback is no improvement above their passband and time domain behaviour doesn't always respond that well. That's something I did in a very small home theater using either two or three subwoofers getting this LF response without smoothing and zero EQ:

However, using subwoofers in both the front and rear and operating in same frequency area really destroys much of the slam feeling. I also try to avoid that.

 

[Bjorn]

Name price and size and do comparative measurements to such a cardioid in the same room and location or send it to Erin, otherwise its just apple vs pears marketing.

As matter of fact you need to optimize for both kinds when it comes to the sub frequencies which is the topic here. That's also why its' so tricky to do a good comparion. As previously mentioned, I plan to do a comparison between cardioid vs monopole subwoofer.

 

[Dmitrij_S]

It's reverberation chamber....Not applicable to a standard living room by any means.

Why not ?! The wave propagation in both a reverberation chamber and a living room is described by the Helmholtz equation. The difference is in the Neumann boundary condition. The walls of the reverberation chamber can be considered as extremely rigid (= has infinite acoustic impedance) so there is no damping of the room modes. While the walls of the living room are more or less flexible and absorptive, i.e. provides some acoustical damping to the room modes. Otherwise, the math of both rooms is the same. Therefore, I do not quite understand the reasoning behind your statements that the results of the article are not applicable to a cardioid in a living room.

 

[Scgorg]

It's the Vera Audio Coherence 12 prototype. Hardly any treatment at all and actually a difficult placement/room (pic below). No sub was used of course.

While I would agree that the D&D 8C/Kii three is not going to deliver the ultimate in-room response (we both know that destructive interference from the floor reflection prevents that), I think it would be most honest to at least post the graphs in the same scale as Erin's if you're going to compare them. Eyeballing it, it appears Erin is using 20-20kHz, 50dB y-axis, and aspect ratio of 25 to 30dB/decade, so that would be a good starting point. Your graphs appear to vary between 80-120dB, and use a much wider frequency span, which of course makes any issues stand out less.

I'm not opposed to this type of discussion in itself, but let's at least try to keep the graph comparisons apples to apples, to the extent possible.

 

[NTK]

Completed the Jupyter notebook describing the theory behind the solution method. Instead of attaching the new ZIP archive here, I added it to post #21, so the previous version and the new version are in the same post. Added 3 more demo cases -- double bass array simulation, comparing it to the "equivalent" single bass array with full rear wall absorption, and cardioid. The notebooks are updated to show a sound pressure response vs time plot at a "listener position". Below are the animations of these new demo cases.

DBA:

SBA with full rear wall absorption as a comparison to DBA:

Cardioid:

 

[Martijn Mensink]

First of all, I apologize for indicating D&D 8C were running cardioid lower in frequency. I should have checked that.

But response is impressive?! No, it's absolutely not. That's a poor response in the area of 100 Hz t0 1000 Hz where the cardioid definetly works. It doesn't matter what Erin writes when the graph show otherwise. Perhaps he has never tested any real good speaker designs and lacks references. I don't know, but that's certainly not a great response or "well behaved room modes above 100 Hz". There's a broad cancellation between 150 Hz and 500 Hz. So not good at all.

I agree. The in-room response in Erin's setup doesn’t look great in the 150–500 Hz range. However, we shouldn’t immediately conclude that cardioid technology isn’t working. The in-room response is not entirely within the speaker's control, other factors are at play.

A good in-room response depends on several things: placement, room acoustics, and listener position, among others. Cardioid dispersion is a tool that can be used to improve in-room response and subjective clarity in this frequency range by reducing rear-wall reflections and controlling early reflections in front of the speaker. But it doesn’t override all acoustic challenges.

Also, while steady-state response is a great predictor of what we hear at low frequencies, it becomes less reliable as frequency increases. Above the modal region, reflections, time-domain behavior, and psychoacoustic factors play a role. That’s why measurements should always be interpreted in context.

If Erin had measured another position or adjusted placement, the results could have been quite different. That’s not an excuse, just a reminder that technology like cardioid dispersion is one piece of the puzzle, not a magic bullet.