• WANTED: Happy members who like to discuss audio and other topics related to our interest. Desire to learn and share knowledge of science required as is 20 years of participation in forums (not all true). There are daily reviews of audio hardware and expert members to help answer your questions. Click here to have your audio equipment measured for free!

Controlled directivity speakers

HooStat

Addicted to Fun and Learning
Joined
May 11, 2020
Messages
808
Likes
841
Location
Calabasas, CA
I was reading @TimVG @Floyd Toole comment about the importance of directivity control as low as possible (see #1,030 ). I had thought that only Dutch & Dutch and Kii were providing that kind of product. But in looking through @pierre's website, I stumbled across the Revel F226Be which seems to have similar directivity control down to about 100 Hz. I have not seen many people comment on this, but maybe I missed it, or maybe I am wrong. I put the horizontal normalized contour plots from @pierre's site below, just for reference. Am I missing something?

The figure for the Revel 226Be on the website was incorrect. The Revel does not have controlled directivity down to 100 Hz as shown. See post #2 below (#2) with the correct plot from the review. The error on @pierre's website has been fixed. Please continue to use this thread as desired to talk about controlled directivity and related topics.

SPL%20Horizontal%20Contour%20Normalized.jpg

SPL%20Horizontal%20Contour%20Normalized.jpg
 
Last edited by a moderator:

Sancus

Major Contributor
Forum Donor
Joined
Nov 30, 2018
Messages
2,653
Likes
6,774
Location
Canada
The chart appears wrong to me. Note in the chart from the actual review, at say 120 degrees and 200hz, the speaker is at -6dB.

In your posted image it shows -12 or -15dB.

That said, larger floorstanders do control directivity better than normal bookshelves, but not as well as the 8C. Which is the cardioid trick, doing it in a small package.

Revel%20F226Be_Horizontal_Spectrogram__Norm_Full.png
 

TimVG

Major Contributor
Joined
Sep 16, 2019
Messages
1,179
Likes
2,527
I would not say with certainty that directivity control as low as possible is needed, but if you can do it down the transition frequency then you can be assured that in the range where it potentially still audibly matters, the reflected sounds are spectrally more similar to the direct sound than with a more conventional loudspeaker, which is never a bad idea.
There's many different ways to achieve a goal like this, but it is tricky if you are trying to do it in a smaller speaker. Not impossible though, as we've already proven. For our R2 project we have settled on controlling the beamwidth down to somewhere between 200 and 300Hz, which should suit many rooms well.
 

Sancus

Major Contributor
Forum Donor
Joined
Nov 30, 2018
Messages
2,653
Likes
6,774
Location
Canada
I would not say with certainty that directivity control as low as possible is needed, but if you can do it down the transition frequency then you can be assured that in the range where it potentially still audibly matters, the reflected sounds are spectrally more similar to the direct sound than with a more conventional loudspeaker, which is never a bad idea.
There's many different ways to achieve a goal like this, but it is tricky if you are trying to do it in a smaller speaker. Not impossible though, as we've already proven. For our R2 project we have settled on controlling the beamwidth down to somewhere between 200 and 300Hz, which should suit many rooms well.
I do wonder if the key to duplicate the magic of the 8C though is to have steep falloff starting already at 60-70 degrees.

Floorstanders do incrementally better here than bookshelves, but it's mostly reduced rearward radiation, at 90 degrees they're usually only down maybe -6dB at most. The 8C is at -15dB by that point already. This reduces interaction to the sides much more than any floorstander I've seen, rather than only to the rear.
 

fineMen

Addicted to Fun and Learning
Joined
Oct 31, 2021
Messages
735
Likes
260
I would not say with certainty that directivity control as low as possible is needed, but ...


I highly appreciate to see this topic discussed individually. I personally, as an 'informed' DIY'er have the opportunity to realise 'cardioid' effortlessly, using DSP and handmade enclosures. I even have all the speakers needed at hand, each in exceptional quality, waveguides included.

So, why do it? The only reason to not do the cardioid is the placement of drivers on the side or back, which kind of looks too technical for my significant other. She would hardly accept a well enclosed, self contained 'box of nothing', but gets for sure annoyed by delicate surfaces all around such a box.

Could You educate me regarding the reasoning behind Your conclusions?
 
Last edited:

dc655321

Major Contributor
Joined
Mar 4, 2018
Messages
1,564
Likes
2,152
The only reason to not do the cardioid is the placement of drivers on the side or back, which kind of looks too technical for my significant other.

She can rest easy: it’s not required to have extra drivers on sides or rear. That is one technique for shaping a cardioid field. Another is passively, as used in Directiva R2 and D&D 8c and other similar designs.
 

TimVG

Major Contributor
Joined
Sep 16, 2019
Messages
1,179
Likes
2,527
I highly appreciate to see this topic discussed individually. I personally, as an 'informed' DIY'er have the opportunity to realise 'cardioid' effortlessly, using DSP and handmade enclosures. I even have all the speakers needed at hand, each in exceptional quality, waveguides included.

So, why do it? The only reason to not do the cardioid is the placement of drivers on the side or back, which kind of looks too technical for my significant other. She would hardly accept a well enclosed, self contained 'box of nothing', but gets for sure annoyed by delicate surfaces all around such a box.

Could You educate me regarding the reasoning behind Your conclusions?

You could hide the drivers easily I suppose with some creative finishing, keeps everyone happy.
Well the reasoning behind it is described in length and detail in Dr. Toole's book. There is an area where the speaker dominates, and one where the room dominates. I suppose if money and complexity is no object you could create a full range constant or controlled directivity loudspeaker. It won't hurt anything, but since the benefit of controlled directivity in the modal range is in the best case 'icing on the cake' and in the worst case 'none at all', we decided, for the time being, to keep the bass module a standard monopole in R2.
 
OP
HooStat

HooStat

Addicted to Fun and Learning
Joined
May 11, 2020
Messages
808
Likes
841
Location
Calabasas, CA
The chart appears wrong to me. Note in the chart from the actual review, at say 120 degrees and 200hz, the speaker is at -6dB.

In your posted image it shows -12 or -15dB.

That said, larger floorstanders do control directivity better than normal bookshelves, but not as well as the 8C. Which is the cardioid trick, doing it in a small package.

Revel%20F226Be_Horizontal_Spectrogram__Norm_Full.png
Good catch -- I didn't check back with the original review. That would explain it. I am sure @pierre will let us know what the issue is when he has time.
 

TimVG

Major Contributor
Joined
Sep 16, 2019
Messages
1,179
Likes
2,527
I do wonder if the key to duplicate the magic of the 8C though is to have steep falloff starting already at 60-70 degrees.

Floorstanders do incrementally better here than bookshelves, but it's mostly reduced rearward radiation, at 90 degrees they're usually only down maybe -6dB at most. The 8C is at -15dB by that point already. This reduces interaction to the sides much more than any floorstander I've seen, rather than only to the rear.

A big horn could perform that aspect even better. The perfect beamwidth however is another discussion and more than likely depends on the room and listening distance, the listener itself, and potentially the source material.
 

Sancus

Major Contributor
Forum Donor
Joined
Nov 30, 2018
Messages
2,653
Likes
6,774
Location
Canada
The perfect beamwidth however is another discussion and more than likely depends on the room and listening distance, the listener itself, and potentially the source material.
I've been wondering for some time if narrower beamwidth in the SBIR range, but then broader beamwidth above that might actually be the ideal for the folks who like to rely on sidewall reflections for envelopment in stereo.

Personally I don't care about that since I think multi-channel is way better, but it does seem possible that something like that might actually be optimal for the "wide directivity in stereo" preference folks.

I don't know of any speaker design like that though.
 

fineMen

Addicted to Fun and Learning
Joined
Oct 31, 2021
Messages
735
Likes
260
You could hide the drivers easily I suppose with some creative finishing, keeps everyone happy.
Well the reasoning behind it is described in length and detail in Dr. Toole's book. ...
I'm quite sure she wouldn't be tricked by any measure of disguise. The front baffle as a delicate surface, covered by non-transparent fancy coloured cloth, o/k-ish, hardly, sadly. But elephant's ears at the sides, or a miraculous back, no deal! ( She cannot wrap her mind around any possible technology, as all too many people of any gender can't. )

Well, I read Floyd's book differently. Didn't he say that reflections were not that much of an issue, as people 'hear through it', to rather take the bigger room, namely from the recording, for real?

One caveat comes to my mind independently. You sparse out reflections with higher directivity. Wouldn't that make the room appear smaller? Means less distributed reflections, leading to deeper nulls and sharper, higher peaks. The equalising property of many independent reflections from all that multiple surfaces is suppressed?

Please report on Your progress!
 

Rick Sykora

Major Contributor
Forum Donor
Joined
Jan 14, 2020
Messages
2,544
Likes
4,766
Location
Stow, Ohio USA
Some of the early results from r2 suggest cabinet depth plays a significant role. I struggle a bit here as I prefer constant directivity (vs controlled). Next it is important to define what is "constant" and understand its relationship to better sound quality. A designer may control directivity in many ways...

Whether what directivity is attained is intended and desirable can be a different matter. In setting requirements for Directiva r2, we had some discussion around directivity and how much better is audibly significant. Like many others, I had not heard a Kii3 or a D&D 8c and so was more constant directivity across more spectrum the answer? I was skeptical. @TimVG helped in this regard as he shared a study. Could not find, but hopefully he can share again.
 
Last edited:

pierre

Addicted to Fun and Learning
Forum Donor
Joined
Jul 1, 2017
Messages
814
Likes
2,262
Location
Switzerland
I was reading @TimVG @Floyd Toole comment about the importance of directivity control as low as possible (see #1,030 ). I had thought that only Dutch & Dutch and Kii were providing that kind of product. But in looking through @pierre's website, I stumbled across the Revel F226Be which seems to have similar directivity control down to about 100 Hz. I have not seen many people comment on this, but maybe I missed it, or maybe I am wrong. I put the horizontal normalized contour plots from @pierre's site below, just for reference. Am I missing something?

SPL%20Horizontal%20Contour%20Normalized.jpg

SPL%20Horizontal%20Contour%20Normalized.jpg

Bad week: 2 bugs! I fixed this one, website is currently uploading corrected contours. Thanks for pointing the discrepancy.
Corrected version for the Revel:

SPL Horizontal Contour Normalized.jpg
 

fluid

Addicted to Fun and Learning
Joined
Apr 19, 2021
Messages
595
Likes
1,017
The only reason to not do the cardioid is the placement of drivers on the side or back, which kind of looks too technical for my significant other.
Significant reduction in output is probably the No1 reason to avoid making a cardioid.
I'm quite sure she wouldn't be tricked by any measure of disguise.
The tightly wrapped fabric sock can hide drivers just as well as it can poor finishing or any number of ills. Some don't appreciate the monolithic appearance though.
 

Attachments

  • IMG_2132.jpg
    IMG_2132.jpg
    146.5 KB · Views: 52

TimVG

Major Contributor
Joined
Sep 16, 2019
Messages
1,179
Likes
2,527
Well, I read Floyd's book differently. Didn't he say that reflections were not that much of an issue, as people 'hear through it', to rather take the bigger room, namely from the recording, for real?

We hear 'through the room' but you still want your reflected sounds to timbrally match the direct sound. The less they match, the more likely we'll interpret those reflected sounds as coloration.

One caveat comes to my mind independently. You sparse out reflections with higher directivity. Wouldn't that make the room appear smaller? Means less distributed reflections, leading to deeper nulls and sharper, higher peaks. The equalising property of many independent reflections from all that multiple surfaces is suppressed?

Having built such a narrow dispersion loudspeaker, I was curious what it would do in a small, treated, room. On the matter of in-room measurements, I did not notice anything unusual in terms of deeper nulls/sharper peaks - but I'm not an acoustician and I didn't look at these things in detail. Since I didn't see any issues, I didn't look into it. The in-room measurements followed the prediction which was a flatter in-room response (with less of a downward trend) due to the directivity properties of the design.

In the end I prefer a somewhat wider cone/dome dispersion profile in a small to normal sized room. R2 is such a design.

In large and very large rooms, the narrower directivity has advantages - this is my personal opinion.
 

TimVG

Major Contributor
Joined
Sep 16, 2019
Messages
1,179
Likes
2,527
@TimVG helped in this regard as he shared a study. Could not find, but hopefully he can share again.

The thesis by Olli Kantamaa can be found here

For those who want the conclusion:

This Thesis has studied the audible effects of increasing the loudspeaker directivity at
low-mid frequencies (LMF). A functional prototype of a compact two–way loudspeaker
having an increased LMF directivity has been designed and implemented. An AB listening
test has been conducted with two loudspeakers A and B, where the LMF directivity was
the only varying characteristic between the loudspeakers A and B. The listening test
results were statistically analysed, showing significant improvements in clarity, absence
of sound colourations, transient reproduction and accuracy of the virtual sound images
improved with increasing LMF directivity. Sound quality impairments were not detected
when the LMF directivity increased. The LMF directivity was the main factor contributing
to the subjective assessment of the sound quality.
Present recommendations (ITU 2015) require increasing the loudspeaker directivity with
increasing frequency. The results of this Thesis do not support such requirement, and
may offer evidence to reconsider the recommendations in this area. It is possible that
the present recommendations stem mainly from the practical characteristics of existing
loudspeaker designs.
 

fineMen

Addicted to Fun and Learning
Joined
Oct 31, 2021
Messages
735
Likes
260
We hear 'through the room' but you still want your reflected sounds to timbrally match the direct sound. The less they match, the more likely we'll interpret those reflected sounds as coloration.
The thesis by Olli Kantamaa can be found here

For those who want the conclusion:
If one take the plunge to read the whole story: "Some of the improved properties may be due changes in the in–situ magnitude response." (page 62) :facepalm:

Why didn't he equialise beforehand? Useless ... and this lack of scientific rigor persists for decades now and renders so much academic effort, that we all pay, inconsequential. Is it on purpose, as to leave some for the generations to come? Science is an eco-system ... too.

But, You've got me :oops:

First I'll test that 'cardiod' thing down-sized to a third, especially targeting it's behaviour close to a (front) wall.

For starters I'll have to prepare best-ever breakfast, and ask my friend for a "go" with a floor-stander showing pumping diaphragms all around its enclosure ... and buy an automated vacuum cleaner additionally. Let's hope for the best.
 
Top Bottom