• 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?

napilopez

Major Contributor
Forum Donor
Joined
Oct 17, 2018
Messages
2,111
Likes
8,433
Location
NYC
I think many if not most of ASR's members are familiar with the Toole/harman school of speaker design - i.e. relatively flat on-axis/listening window and smoothly changing or constant directivity off-axis - and that research shows this approach is highly correlated with listener preference. It's fair to assume the average speaker engineer knows this. What then, are the reasons so many speaker manufacturers seem to deviate from this approach?

The cynical answer - and probably truthful, in some cases - is that it's just aesthetics and marketing. Some speaker designs look nice or unique, and they might seem more impressive when listened for a couple of minutes in isolation from other speakers, especially when playing sparse music that is unlikely to reveal flaws.

Some people seem to worry that if all speakers followed the Toole approach, all speakers would look and sound the same, although I think @Ilkless made a good case against that in this post. Of course, there also budget constraints on the low end of the price scale - I don't mean to trivialize designing a technically good loudspeaker - but there are certainly affordable speakers that perform admirably as well.

But it seems some manufacturers simply don't 'agree' with the research, or at least actively choose to go against it with their own designs. Sometimes I see measurements that are completely puzzling, but I have to assume the engineers have their reasons. What do you think? Are there any example where a speaker appears to go against the research but ends up sounding good in-room?
 
Last edited:

sergeauckland

Major Contributor
Forum Donor
Joined
Mar 16, 2016
Messages
3,440
Likes
9,100
Location
Suffolk UK
I see several reasons for manufacturers not achieving a flat on-axis response.
Firstly, they may lack the technical measuring equipment to do so. In the past, measuring loudspeakers involved renting an anechoic chamber and/or open air measurements on a tall gantry, both expensive and inconvenient. With current software, even free, this is no longer a credible reason.

Secondly, as suggested above, they may have a marketing reason for wanting to sound 'different' on dealer demos, and thirdly, the ego of the designer who claims 'golden ears' and consequently thinks they can 'voice' their loudspeakers to sound better than a flat response.

These days, there's certainly no reason why a loudspeakers shouldn't be flat unless deliberately designed not to be. I really do wonder what's gone on in the mind of designers of dreadful stuff like Zu or Boenicke.

S
 

RayDunzl

Grand Contributor
Central Scrutinizer
Joined
Mar 9, 2016
Messages
13,201
Likes
16,983
Location
Riverview FL
Are there any example where a speaker appears to go against the research but ends up sounding good in-room?


I like my MartinLogan reQuests, the design of which which rates poorly on the Harman/Toole/Olive preference scales, but after 21 years with them, and despite having a set of "approved" speakers right next to them for easy comparison and daily casual use, don't feel a need to take them to the dump.
 
Last edited:
OP
N

napilopez

Major Contributor
Forum Donor
Joined
Oct 17, 2018
Messages
2,111
Likes
8,433
Location
NYC
These days, there's certainly no reason why a loudspeakers shouldn't be flat unless deliberately designed not to be. I really do wonder what's gone on in the mind of designers of dreadful stuff like Zu or Boenicke.

Such mysteries are the kind I'm trying to puzzle out! I haven't heard those specific brands, but the measurements sure don't make me want to when there are so many other speakers that appear to be more 'tastefully designed.'

I like my MartinLogan reQuests, which rate poorly on the Harman/Toole/Olive preference scales, but after 21 years with them, and despite having a set of "approved" speakers right next to them for easy comparison and daily casual use, don't feel a need to take them to the dump.

Yeah, I remember you saying that =] A lot of people like their B&W's too, even though some of their most recent models seem to measure strangely.
 

daftcombo

Major Contributor
Forum Donor
Joined
Feb 5, 2019
Messages
3,687
Likes
4,068
Perhaps people want to have the speakers pointing straight fot aesthetical reasons? In that case the flat-axis response might be felt as a lack of treble?
 
OP
N

napilopez

Major Contributor
Forum Donor
Joined
Oct 17, 2018
Messages
2,111
Likes
8,433
Location
NYC
Perhaps people want to have the speakers pointing straight fot aesthetical reasons? In that case the flat-axis response might be felt as a lack of treble?

Yeah, that's why I mentioned the listening window as well, as it reflects a wider set of use cases. My post is more about addressubg speakers with notable dips and peaks in the frequency response.

But to your point, indeed, I prefer the look of speakers pointing straight forward in my living room, though so far only the KEF speakers I've tried have sounded best this way. Both the R3 and LS50W I've tested sounded and measured best about 15 degrees off-axis.
 

daftcombo

Major Contributor
Forum Donor
Joined
Feb 5, 2019
Messages
3,687
Likes
4,068
Here is a funny read which might answer your question then:

Dude, Flat sound like ass.
Dude, some people prefer color in their sound. I've got a menu for that, to add color back into the corrections for that transparency, speed, attack, clarity, etc. that audio reviewers go on about. These are the secrets that I've unlocked so far. I have not found them to universally work for all speakers. Use at your own risk, goofy sounds can be addictive. Also remember to use the preamp level to cut it make sure there is no gain added to the digital signal, digital distortion is real and it's ugly.


Move imaging forwardFilter 1: ON PK Fc 1600 Hz Gain 2.0 dB Q 1.00
Move imaging backwardFilter 1: ON PK Fc 1600 Hz Gain -2.0 dB Q 1.00
Transparency(or the B&W filter)Filter 1: ON PK Fc 5000 Hz Gain -5.0 dB Q 4.00
WarmthFilter 1: ON PK Fc 220 Hz Gain 2.0 dB Q 1.50
DetailFilter 1: ON PK Fc 8000 Hz Gain 2.0 dB Q 2.50
AnalyticalFilter 1: ON PK Fc 18000 Hz Gain 4.0 dB Q 0.70

Source: http://noaudiophile.com/DSP_Corrections/
 
OP
N

napilopez

Major Contributor
Forum Donor
Joined
Oct 17, 2018
Messages
2,111
Likes
8,433
Location
NYC
I don't understand the relation with the window. Can you explain?

The Listening Window is measured differently by different folks, but for Harman, for example, it's an average of the on-axis response and the, +/- 10/20/30 horizontal, and +/- 10 vertical off axis response. So the listening window roughly accounts for speakers that aren't toed in all the way/are pointing straight out. I would imagine in most homes, speakers would be positioned such that the main listening seat is no more than 30 degrees off axis. That would be the case if the speaker positioning is an equilateral triangle but the speakers were positioned straight out.

Here is a funny read which might answer your question then:
...
Source: http://noaudiophile.com/DSP_Corrections/

Don't get me wrong, I definitely think there is some room for preference, and I liked to a post in my OP that showed how a variety of different designs can still follow modern audio science. I don't want all speakers to sound exactly the same. It's more about hitting a certain quality standard, I guess.
 
Last edited:

daftcombo

Major Contributor
Forum Donor
Joined
Feb 5, 2019
Messages
3,687
Likes
4,068
I get you on the window.

Try those EQ if you have some time.
The guy definitely nailed it.
The Detail and the Tranparency one are particularly entertaining.
 

Rja4000

Major Contributor
Forum Donor
Joined
May 31, 2019
Messages
2,682
Likes
4,220
Location
Liège, Belgium
Hi
I have a few loudspeaker pairs at home.
A pair of Kef LS50, Focal Aria 906, Focal Aria 948, Cabasse Skiff and Genelec 1032A
I use Dirac Live on the 948 and Roon REW-measured convolution filters on the others (+/-2dB for the later), all set for Harman -0.9dB/octave slope, with some raise after 10kHz (I'm getting old).
And still, they all sound different.

On top of the frequency response, a lot of other factors 'makes' the loudspeaker sound.
Dynamic response, response around the crossover frequency, 3D directivity, bass extension, room interaction of cours (and even the look, probably). Not to mention the amplifier and source chain behind them.

There is no 'perfect speaker'. At least not in the price range (or size or environment) I can afford.
It's all about compromises. And each compromise will add it's own signature to the speaker 'voice'.
That's also why you can typically recognize a speaker family 'signature', as I do, as an example, between the small-ish Focal Aria 906 and the huge 948. Or as you may find between Genelec speakers.

And this, so far, doesn't measure well...
At least, as far as I know.
 
Last edited:

amirm

Founder/Admin
Staff Member
CFO (Chief Fun Officer)
Joined
Feb 13, 2016
Messages
44,374
Likes
234,451
Location
Seattle Area
But it seems some manufacturers simply don't 'agree' with the research, or at least actively choose to go against it with their own designs.
The #1 thing in a successful speaker business is marketing. Unfortunately consumer awareness of Dr. Toole research is close to zero. I am actually pleasantly surprised how much our members know about it. In other fora people either don't know what it is or hate it on principal or both. Without consumer pull, marketing departments are not going to push for compliance with this research, or highlighting it in their marketing.
 

Ron Texas

Master Contributor
Joined
Jun 10, 2018
Messages
6,078
Likes
8,914
The #1 thing in a successful speaker business is marketing. Unfortunately consumer awareness of Dr. Toole research is close to zero. I am actually pleasantly surprised how much our members know about it. In other fora people either don't know what it is or hate it on principal or both. Without consumer pull, marketing departments are not going to push for compliance with this research, or highlighting it in their marketing.

From someone in the business. Marketing is the difference between staying in business and going bust.
 
OP
N

napilopez

Major Contributor
Forum Donor
Joined
Oct 17, 2018
Messages
2,111
Likes
8,433
Location
NYC
The #1 thing in a successful speaker business is marketing. Unfortunately consumer awareness of Dr. Toole research is close to zero. I am actually pleasantly surprised how much our members know about it. In other fora people either don't know what it is or hate it on principal or both. Without consumer pull, marketing departments are not going to push for compliance with this research, or highlighting it in their marketing.

I always have to remind myself to not underestimate the power of marketing.

It's a shame so few people are aware of the research, but maybe, just maybe things are changing. You see a lot of people talking about measurements on younger-leaning platforms like reddit nowadays. And I'm hoping I can build some mainstream awareness through my corner of the tech media.
 

amirm

Founder/Admin
Staff Member
CFO (Chief Fun Officer)
Joined
Feb 13, 2016
Messages
44,374
Likes
234,451
Location
Seattle Area
It's a shame so few people are aware of the research, but maybe, just maybe things are changing. You see a lot of people talking about measurements on younger-leaning platforms like reddit nowadays. And I'm hoping I can build some mainstream awareness through my corner of the tech media.
I know how to solve this problem but unfortunately it costs $60,000 in test equipment to generate the "spinorama" measurements. We have done it with measurements of audio electronics. We can do the same with speaker measurements. That cost (for electronics testing) was around $30,000. It is a struggle for me to make the investment in speaker testing but I am still considering it.
 

MZKM

Major Contributor
Forum Donor
Joined
Dec 1, 2018
Messages
4,244
Likes
11,472
Location
Land O’ Lakes, FL
1) Research by Sean Olive and company have shown that the average preference is for decently more bass and a bit less treble, so dead neutral (anechoically) is actually not ideal, it‘s just close, and if the off-axis is well-controlled, it can be easily EQ’d.

2) Some speakers are designed for no toe-in or other special arraignments. The HSU CCB-8 for instance is designed to be toed-in extremely, 15° past the listener:
image

So that it has a similar effect to what Ohm Walsh speakers do:
frs.875x0-is.jpg

And thus they were designed to be pretty bright if listened on-axis.

3) They may not be able to design a speaker/driver that has good off-axis performance. I often see speakers that have peaks/dips off-axis and thus the designer added dips/peaks on-axis to compensate, but otherwise the off-axis is good.

4) For those mentioning dipoles, you can’t really measure their FR anechoically, they just look terrible, doing in-room multi-seat averaged responses (which Stereophile does) is more telling.

5) They are designed for special use cases. Zu speakers measure very poorly, but are very high sensitivity because they are aimed at people using 2W tube amps, and sometimes their impedance curves allows them to have a better FR than measurements show (their FR will be altered to follow their impedance curve).

6) Some companies are just incompetent and are more focused on visual appeal.
 
Last edited:

Kal Rubinson

Master Contributor
Industry Insider
Forum Donor
Joined
Mar 23, 2016
Messages
5,273
Likes
9,790
Location
NYC
2) Some speakers are designed for no toe-in or other special arraignments. The HSU CCB-8 for instance is designed to be toed-in extremely, 15° past the listener:
image
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.
 
OP
N

napilopez

Major Contributor
Forum Donor
Joined
Oct 17, 2018
Messages
2,111
Likes
8,433
Location
NYC
Good stuff. I've read about the HSU's extreme toe-in, and I believe Geddes also advocates for similar extreme toe-in to help widen the sweet spot. It's an interesting idea that I've had good results with on some speakers, though aesethetically it tends to bothers me a fair bit. I had never considered the impedance curve and tubes re:zu speakers, that's an interesting point. This interview with Zu's Sean Casey and JA is an interesting watch.
 

Blumlein 88

Grand Contributor
Forum Donor
Joined
Feb 23, 2016
Messages
20,524
Likes
37,057
The #1 thing in a successful speaker business is marketing. Unfortunately consumer awareness of Dr. Toole research is close to zero. I am actually pleasantly surprised how much our members know about it. In other fora people either don't know what it is or hate it on principal or both. Without consumer pull, marketing departments are not going to push for compliance with this research, or highlighting it in their marketing.
I love the part of the story where Harman's own engineers first took sighted test and blinded. In sighted they of course thought highly of their own speakers and exotics. Without retelling it all, this one thing should indicate designers elsewhere that don't do blind reality checks are necessarily designing partly by predjudice and bias. It has to hamper their designs.
 

Blumlein 88

Grand Contributor
Forum Donor
Joined
Feb 23, 2016
Messages
20,524
Likes
37,057
snippage....

4) For those mentioning dipoles, you can’t really measure their FR anechoically, they just look terrible, doing in-room multi-seat averaged responses (which Stereophile does) is more telling.

5) They are designed for special use cases. Zu speakers measure very poorly, but are very high sensitivity because they are aimed at people using 2W tube amps, and sometimes their impedance curves allows them to have a better FR than measurements show (their FR will be altered to follow their impedance curve).

6) Some companies are just incompetent and are more focused on visual appeal.

Apparently in blind Harman listening tests dipoles do about as well as their anechoic measurements would suggest. I also have applied some curves from them to flatter measuring speakers. They don't of course sound like the dipoles then, but they are more than halfway there in the prime listening position.

ESL's also sometimes benefit from the output impedance of the higher powered push-pull tube units. It can give a decibel or two more low end over a couple octaves and drops about that much in the treble.
 
Top Bottom