• WANTED: Happy members who like to discuss audio and other topics related to our interest. Desire to learn and share knowledge of science required. There are many reviews of audio hardware and expert members to help answer your questions. Click here to have your audio equipment measured for free!

Why don't all speaker manufacturers design for flat on-axis and smooth off-axis?

Daverz

Major Contributor
Joined
Mar 17, 2019
Messages
1,294
Likes
1,451
This is fundamentally what was recommended by Blumlien. It compensated for small head/position changes since a move in either direction brought one closer to the ipsilateral speaker while simultaneously closer to on-axis for the contralateral one.

You just wanted to write ipsilateral in a post!
 

FrantzM

Major Contributor
Forum Donor
Joined
Mar 12, 2016
Messages
4,335
Likes
7,719
While I respect the Toole/Harman School of speaker design, I would not go as far as thinking it is the only valid way to design a speaker. There are other schools/philosophy/design concepts. People like Geddes take a different approach. Granted he did/doesn't have the vast Harman resources but his research is solid and backed by science and empirical results. The people at places like B&O and Kef do not seem to follow the same principles either.
I do believe the Harman/Toole is a valid way. There are others.
 
Last edited:

Blumlein 88

Grand Contributor
Forum Donor
Joined
Feb 23, 2016
Messages
20,521
Likes
37,050
While I respect the Toole/Harman School of speaker design, I would not go as far as thinking it is the only valid way to design a speaker. There are other schools/philosophy/design concepts. People like Geddes take a different approach. Granted he did/doesn't have the vast Harman resources but his research is solid and backed by science and empirical results. The people at places like B&O and Kef do not seem to follow the same principles abide by the same principles either.
I do believe the Harman/Toole is a valid way. There are many

Well we have to break the circle somewhere. The more of their designs I hear, the more I'm coming around to the idea that in totality their method has not been bettered by others. There may be this or that which something will be better than Harman, but the testing by listening has put them way ahead of most. I'm tending to think anything that will better them will do so by scoring well in the all the ways they design for and something beyond that. I'd like to know how a blind Harman facility test of the kii Three or Dutch&Dutch 8's would go vs the top Revel or JBL M2.

Now does that mean only their speakers are competitive? No, some others are doing similar designs. And there is always the effect of seeing what you are listening to. Maybe other good speakers score well, but not as well as some Revel or JBL, and yet by the visual and tactile and all other factors generate considerable satisfaction and a predisposition to judge as sounding better the speaker in use. Even long term that can leave a music lover feeling better than with a Harman design.
 

amirm

Founder/Admin
Staff Member
CFO (Chief Fun Officer)
Joined
Feb 13, 2016
Messages
44,368
Likes
234,381
Location
Seattle Area
The people at places like B&O and Kef do not seem to follow the same principles abide by the same principles either.
KEF does actually. For example: https://us.kef.com/explore-kef/all-about-sound/kefs-audio-reproduction-philosophy

1563251527576.png


If you read their white paper on coaxial driver, there is strong coverage of off-axis response being similar to on axis: http://www.kef.com/uploads/files/THE_REFERENCE/REF_White_Paper_preview_path_200514.pdf

Including this key reference:

1563251879526.png


Many companies have adopted this strategy but most are not coming right out and screaming "as researched by Dr. Toole." Indeed few companies dare to argue against the goodness of direct and indirect sounds having similar tonality.
 

March Audio

Master Contributor
Audio Company
Joined
Mar 1, 2016
Messages
6,378
Likes
9,317
Location
Albany Western Australia
While I respect the Toole/Harman School of speaker design, I would not go as far as thinking it is the only valid way to design a speaker. There are other schools/philosophy/design concepts. People like Geddes take a different approach. Granted he did/doesn't have the vast Harman resources but his research is solid and backed by science and empirical results. The people at places like B&O and Kef do not seem to follow the same principles abide by the same principles either.
I do believe the Harman/Toole is a valid way. There are many
As Amir said KEF do.

See this lecture from 1:06 and the comments of the KEF guy.

 
OP
N

napilopez

Major Contributor
Forum Donor
Joined
Oct 17, 2018
Messages
2,109
Likes
8,420
Location
NYC
While I respect the Toole/Harman School of speaker design, I would not go as far as thinking it is the only valid way to design a speaker. There are other schools/philosophy/design concepts. People like Geddes take a different approach. Granted he did/doesn't have the vast Harman resources but his research is solid and backed by science and empirical results. The people at places like B&O and Kef do not seem to follow the same principles abide by the same principles either.
I do believe the Harman/Toole is a valid way. There are many

I certainly think there's room for flexibility, and it would be really interesting if more research emerged to contradict some of the Harmanesque findings, but I wouldn't consider the examples you provided as being particularly different from Harman - more like variations on a theme. KEF pretty explicitly says its views line up with Toole's research in its Reference series whitepaper (Section 4.1.5), at least with regards to frequency response on and off-axis. It even provides spinorama data for a Reference 5 prototype, and it provides spin-ish graphs for its new R series speakers as well :):

Snag_20b3300c.png

KEF Reference 5 Prototype Spins:
Snag_20b6c7b9.png


For Bang and Olufsen, I've only really looked into their flagship, the Beolab 90, but that one at least seems to be pretty harmanesque too. It too has a whitepaper available, which is pretty massive, but reveals much of the design philosophy. Though there aren't frequency plots, it does say B&O performs some tuning to balance the on-axis and power response in room but the speaker actually includes a 'flat on-axis' mode in the software should you prefer it. B&O does show polar plots for the different directivity modes; the narrow and wide modes look pretty smooth.

It appears to be a ridiculously customizable speaker - not many speakers allow you to shape both frequency response and directivity. I'd love to see more speakers like this.

And my impression has been that Gedess generally agree with Toole in the grand scheme of things. The main thing I recall is he advocates for narrow directivity and Floyd prefers wide, but these are matters of preference

The speakers I'm referring to in my OP are more like the aforementioned Zus, some B&Ws, Devores, etc. In particular more expensive models where budget is presumably not a concern. Sometimes these speakers display measurements that are well far from flat on-axis and/or display some wonky dispersion behavior. I really am not passing judgement on these brands or individual speakers as I haven't heard all of them, but I'm genuinely curious as to the potential mindset behind their designs.

Also note that I'm not saying Harman makes the best speakers (I haven't even heard any of the Revel speakers, arguably its most lauded brand). I'm just broadly referring to the basic principles of flattish on-axis axis and smoothly changing or constant directivity off-axis that its research has backed up. Different brands will have different philosophies and will execute differently on just how flat 'flat on-axis' needs to be, the importance of reducing distortion, wide vs narrow dispersion, efficiency, etc.

EDIT: I now see two others have mentioned the KEF bit. Woops, that's what I get for not refreshing =]
 

Blumlein 88

Grand Contributor
Forum Donor
Joined
Feb 23, 2016
Messages
20,521
Likes
37,050
The 1st order crossover seems a bad idea to me. I get the adoration for phase coherence as it is marketed. But that only applies to one plane, necessarily ruins off axis response in other directions, and means heroic drivers are needed to make this come close to working.

I wonder how much better some of Jim Thiel's speakers might have been if he had given up the 1st order crossovers for some 4th order crossovers.

Thiel is gone, but we still have Vandersteen among others promoting and having built their rep on the mystique of 1st order crossovers.

Vandersteen Treo lateral response from Stereophile measures:
1563256948042.png


And vertical response:
1563256965878.png


It does have this nice step response, if you measure it in just the right plane.
1563257057106.png


I suppose outdoors or in an anechoic chamber on the correct plane for the listening position this speaker would be great.
 
Last edited:

Ilkless

Major Contributor
Forum Donor
Joined
Jan 26, 2019
Messages
1,756
Likes
3,437
Location
Singapore
For Bang and Olufsen, I've only really looked into their flagship, the Beolab 90, but that one at least seems to be pretty harmanesque too. It too has a whitepaper available, which is pretty massive, but reveals much of the design philosophy. Though there aren't frequency plots, it does say B&O performs some tuning to balance the on-axis and power response in room but the speaker actually includes a 'flat on-axis' mode in the software should you prefer it. B&O does show polar plots for the different directivity modes; the narrow and wide modes look pretty smooth.

It appears to be a ridiculously customizable speaker - not many speakers allow you to shape both frequency response and directivity. I'd love to see more speakers like this.

B&O's "tonmeister" Geoff Martin seems to be more than conversant with the importance of smooth directivity.
 

Thomas_A

Major Contributor
Forum Donor
Joined
Jun 20, 2019
Messages
3,421
Likes
2,406
Location
Sweden
As to the question in the thread title I would add rhe following: why are there so few loudspeaker that are deisgned for specific room sizes and placements in the room? A few ones allow for corrections in frequency response and very few allow for correction of directicity pattern depending on placment and room size. Also there are none that corrects for sterep vs multichannel diffeences?
 

Krunok

Major Contributor
Joined
Mar 25, 2018
Messages
4,600
Likes
3,065
Location
Zg, Cro
A few ones allow for corrections in frequency response and very few allow for correction of directicity pattern depending on placment and room size.

I know how to correct on-axis FR but how exactly would you correct for bad directivity pattern?
 

Thomas_A

Major Contributor
Forum Donor
Joined
Jun 20, 2019
Messages
3,421
Likes
2,406
Location
Sweden
I know how to correct on-axis FR but how exactly would you correct for bad directivity pattern?

Beolab 90 or similar.
 

Thomas_A

Major Contributor
Forum Donor
Joined
Jun 20, 2019
Messages
3,421
Likes
2,406
Location
Sweden

Thomas_A

Major Contributor
Forum Donor
Joined
Jun 20, 2019
Messages
3,421
Likes
2,406
Location
Sweden
And how would you correct that?

With a slight increase in the 1-2 kHz region and a dip in the 3-4 and 7 kHz region.
 

Krunok

Major Contributor
Joined
Mar 25, 2018
Messages
4,600
Likes
3,065
Location
Zg, Cro
With a slight increase in the 1-2 kHz region and a dip in the 3-4 and 7 kHz region.

Wouldn't you say that a correction like that one should be implemented in a mutichannel sound processors when working with phantom centre to increase ability to better play speech? Why do you think it has anything to do with speakers?
 

Thomas_A

Major Contributor
Forum Donor
Joined
Jun 20, 2019
Messages
3,421
Likes
2,406
Location
Sweden
Wouldn't you say that a correction like that one should be implemented in a mutichannel sound processors when working with phantom centre to increase ability to better play speech? Why do you think it has anything to do with speakers?

Well it is part of the inherent flaw of stereo reproduction. So far I have not seen any such corrections when playing stereo recordings via AV recievers. There is one commercial 3.x system playing stereo recordings using three speakers that I know of that probably do some of these corrections.
 
Top Bottom