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

[Cosmik]

Isn't Harman's spinorama confirmed in multiple tests to reflect what we hear and what we like?

The problems with slim floorstanders that RE Greene identifies don't contradict the need for smooth off-axis response. Given two slim floorstanders, the one with the the smoother off-axis response (all else being equal) will no doubt be preferred. But smooth off-axis response isn't the whole story. Where the baffle step lies is another factor, for example.

In listening tests, as RE Greene points out, even professionals get used to a particular sound. This is one element of the "circle of confusion". As he points out, "detail" or "transparency" can be seductive but wrong. It starts off sounding good, perhaps, but over time, the sound wears you down and you go out looking for your next speaker. I don't see why the Harman tests should be immune from this, and nor do they address all rooms, all speaker placements, all types of music, all volume settings, etc.

RE Greene is trying to circumvent all that by suggesting a reason why all the FR measurements, target curves etc. never seem to add up to a consistent story. It's necessary to do it by reasoning alone, not random trials and feedback from measurements. And that's what the Grimm people were doing, too.

 

[Bjorn]

From the Electromotion ESL manual.. Nothing specatular about positioning there.

View attachment 29620

So? Making arguments based on a marketing brochure or what most at audio shows do is futile. Studies shows that dipoles work generally best with a certain position and which is quite different from monopoles especially in the lows. But it will also depend on the room and you have to take into consideration the front wall contribution. Point is that you can get very different results here, something that a serious study should take into consideration IMO.

 

[Krunok]

The problems with slim floorstanders that RE Greene identifies don't contradict the need for smooth off-axis response. Given two slim floorstanders, the one with the the smoother off-axis response (all else being equal) will no doubt be preferred. But smooth off-axis response isn't the whole story. Where the baffle step lies is another factor, for example.

In listening tests, as RE Greene points out, even professionals get used to a particular sound. This is one element of the "circle of confusion". As he points out, "detail" or "transparency" can be seductive but wrong. It starts off sounding good, perhaps, but over time, the sound wears you down and you go out looking for your next speaker. I don't see why the Harman tests should be immune from this, and nor do they address all rooms, all speaker placements, all types of music, all volume settings, etc.

RE Greene is trying to circumvent all that by suggesting a reason why all the FR measurements, target curves etc. never seem to add up to a consistent story. It's necessary to do it by reasoning alone, not random trials and feedback from measurements. And that's what the Grimm people were doing, too.

Are there any measurements or studies confirming this "slim baffle = bad sound" theory?

How do you explain that Ultima2 easilly won blind test over M2 although it is much "slimmer"?

And finally, when I look at some serious flagship speakers, not only from Harman but from Kef, Dynaudio, Hedd.. they all appear slim to me. Do you really think RE Greene knows something that development guys in these companies don't?

 

[Krunok]

So? Making arguments based on a marketing brochure or what most at audio shows do is futile. Studies shows that dipoles work generally best with a certain position and which is quite different from monopoles especially in the lows. But it will also depend on the room and you have to take into consideration the front wall contribution. Point is that you can get very different results here, something that a serious study should take into consideration IMO.

Which studies? And why isn't ML suggesting their customers to position their speakers according to those studies?

MLs are hybrids where panel is sitting on top of SW so your argument about "lows" doesn't really apply. Dipoles without classic sub don't have true bass to speak about anyway.

P.S. User manuals are not written by marketing but rather technical support people. Product info brochures are written by marketing.

 

[Cosmik]

Are there any measurements or studies confirming this "slim baffle = bad sound" theory?

How do you explain that Ultima2 easilly won blind test over M2 although it is much "slimmer"?

And finally, when I look at some serious flagship speakers, not only from Harman but from Kef, Dynaudio, Hedd.. they all appear slim to me. Do you really think RE Greene knows something that development guys in these companies don't?

Yes, I think he does.

As he describes, when it comes to preference-based listening tests, people cannot separate what they are used to from what is right; they can only judge which speakers they are presented with; they can only judge a small subset of music in a small subset of rooms, volume levels; they cannot account for a novelty factor that wears off after the experiment is over etc. etc.

Just because 'data' exists, doesn't mean its origins or interpretation aren't flawed.

A slim speaker is objectively further from neutral (if neutral would be uniform dispersion angle at all frequencies) than a wider speaker.

 

[Krunok]

Which seems to be why the Grimm people, for example, created a different speaker;

They certainly look interesting but I wasn't able to find any objective review supported by measurements of Grimm LS1 speakers. Do you have any?

 

[Krunok]

Yes, I think he does.

As he describes, when it comes to preference-based listening tests, people cannot separate what they are used to from what is right; they can only judge which speakers they are presented with; they can only judge a small subset of music in a small subset of rooms, volume levels; they cannot account for a novelty factor that wears off after the experiment is over etc. etc.

Just because 'data' exists, doesn't mean its origins or interpretation aren't flawed.

The point is that a slim speaker is further from neutral (i.e. neutral would be uniform dispersion angle at all frequencies) than a wider speaker.

So there is Greene with his "the wider the baffle the better the sound" theory but no studies, measurements nor blind tests supporting this theory?

 

[Cosmik]

So there is Greene with his "the wider the baffle the better the sound" theory but no studies, measurements nor blind tests supporting this theory?

It's very difficult to get across the idea that 'data' isn't necessarily 'truth'! Sometimes it just has to be the development of ideas alone. Preference-based listening tests are open to so many problems that they can't be relied upon.

The notion that a listener can be trained to produce 'truth' is not a fact. The only thing that can be proved about trained listeners is that they are consistent, perhaps - a circular, self-referencing, bootstrapping measure against themselves. Consistency does not mean ability to discern truth/quality/meaningfulness etc.

 

[Krunok]

It's very difficult to get across the idea that 'data' isn't necessarily 'truth'! Sometimes it just has to be the development of ideas alone. Preference-based listening tests are open to so many problems that they can't be relied upon.

The notion that a listener can be trained to produce 'truth' is not a fact. The only thing that can be proved about trained listeners is that they are consistent, perhaps - a circular, self-referencing, bootstrapped measure against themselves. Consistency does not mean ability to discern truth/quality/meaningfulness etc.

I apologise if this will sound harsh but to me it seems that with this "philosophy" you are trying to compensate for the lack of supporting studies, blind tests and measurements.

 

[Cosmik]

I apologise if this will sound harsh but to me it seems that with this "philosophy" you are trying to compensate for the lack of supporting studies, blind tests and measurements.

I'm not in the slightest embarrassed about making the argument. I don't know whether anyone has ever tested to see whether an amplifier should invert phase or whether a DAC should be monotonic, or whether an amplifier shouldn't compress loud transients. In all cases, I am confident that a definition without reference to empirical studies is best.

The case with speakers is a little more complex to define, and existing folklore confuses the situation, but the same principle applies: the speaker's function can, and should, be defined without reference to empirical studies - as much as possible. Empirical studies would just produce 'noise'.

 

[Krunok]

The case with speakers is a little more complex to define, and existing folklore confuses the situation, but the same principle applies: the speaker's function can, and should, be defined without reference to empirical studies. Empirical studies would just produce 'noise'.

Ok. With amps and DACs you certainly don't need empirical studies to confirm they are functioning well - measurements can do that job just fine. Are you saying the same principle should be applied to speakers?

 

[Cosmik]

Ok. With amps and DACs you certainly don't need empirical studies to confirm they are functioning well - measurements can do that job just fine. Are you saying the same principle should be applied to speakers?

Yes, basically. But with an acceptance that complete neutrality isn't possible, and there are different flavours of neutral e.g. omnidirectional speakers can be neutral just as more directional speakers can. The potential customer still has to decide on that aspect, I would say - the Grimm paper talks about the "ghostly brass section" and recorded audio as an "ill-posed problem". An analogy might be choosing whether to buy a 42" TV or 100". They're both neutral, but sometimes suited to different things. But this doesn't mean that "anything goes" and that empirical testing might reveal a hitherto unknown type of television set that viewers prefer because of its heart-shaped screen and pink colour cast.

 

[Krunok]

Yes, basically. But with an acceptance that complete neutrality isn't possible, and there are different flavours of neutral e.g. omnidirectional speakers can be neutral just as more directional speakers can.

But isn't that the argument in favor of checking with a series of blind tests if your measurement technique correlates with the customers preferred choice? Isn't that exactly how Harman developed spinorama?

 

[Cosmik]

But isn't that the argument in favor of checking with a series of blind tests if your measurement technique correlates with the customers preferred choice? Isn't that exactly how Harman developed spinorama?

I'm suggesting that neutral speakers are rigidly defined, just as a TV is defined as having uniformly-spaced dots placed in a rigid matrix providing an aspect ratio of 16:9, or whatever. No TV is perfect, of course, and a TV manufacturer might devise a machine to measure the accuracy of their TVs.

The Spin-o-rama is the speaker equivalent: it measures the closeness to neutrality that the speaker achieves without requiring a human to listen to them. If the manufacturer wants to bring in some humans for fun and to confirm how great their speakers are, that's fine, but don't expect them to discover a new paradigm.

 

[Krunok]

I'm suggesting that neutral speakers are rigidly defined, just as a TV is defined as having uniformly-spaced dots placed in a rigid matrix providing an aspect ratio of 16:9, or whatever. No TV is perfect, of course, and a TV manufacturer might devise a machine to measure the accuracy of their TVs.

The Spin-o-rama is the speaker equivalent: it measures the closeness to neutrality that the speaker achieves without requiring a human to listen to them. If the manufacturer wants to bring in some humans for fun and to confirm how great their speakers are, that's fine, but don't expect them to discover a new paradigm.

My opinion is that speakers are technically not perfect enough for that kind of approach so designers/testers should check with people what kind of sound they prefer.

 

[Juhazi]

R.E. Greene is a rather controversial person in audio journalism. I can share many of his opinions, but strongly disagree with some. http://www.regonaudio.com/ He writes regularly to TAS website https://www.theabsolutesound.com/

I can't understand his dislike of narrow cabinets. Eg. baffle loss/step is a very simple and easily handled phenomenom well know for ages - but still ignored by some "purists" who like eg. fullrange speakers. Edge contour and tweeter positioning on the baffle can be handled too as well in narrow and wide cabs
http://www.mh-audio.nl/Calculators/BDL.html

Narrow cabinets sell better, because they don't look so big and heavy.
A good speaker with flat medium width 14" cabinet https://www.stereophile.com/content/magico-s5-mkii-loudspeaker-measurements
A good speaker with narrow baffle https://www.stereophile.com/content/vivid-audio-giya-g3-loudspeaker-measurements

Pic of Giya added -designed by the same guy who created B&W Nautilus!

 

[Krunok]

A good speaker with narrow baffle https://www.stereophile.com/content/vivid-audio-giya-g3-loudspeaker-measurements

Measurements do look good but with that fat 34cm wide belly they are not really slim - they actually look fat.

 

[napilopez]

Early reflection curve is the average of eight responses: ±40°, ±60° & ±80° in the horizontal plane and ±50° in the vertical plane.

I've seen the above mentioned elsewhere, but in section 5.3 of Toole's book the early reflections curve is actually characterized as an average of:

• Floor bounce: average of 20 °, 30 °, 40 ° down
• Ceiling bounce: average of 40 °, 50 °, 60 ° up
• Side wall bounces: average of ± 40 °, ± 50 °, ± 60 °, ± 70 °, ± 80 ° horizontal
• Front wall bounce: average of 0 °, ± 10 °, ± 20 °, ± 30 ° horizontal
• Rear wall bounces: average of 180 °, ± 90 ° horizontal

That includes a lot more vertical data. So I'm wondering if I'm missing something - are there two standards for the curve?

So? Making arguments based on a marketing brochure or what most at audio shows do is futile. Studies shows that dipoles work generally best with a certain position and which is quite different from monopoles especially in the lows. But it will also depend on the room and you have to take into consideration the front wall contribution. Point is that you can get very different results here, something that a serious study should take into consideration IMO.

As Krunok pointed out, manuals are usually written by technically capable people with input from the engineers, as far as I know.

In any case, if a company can't tell you how to position speakers that require super special positioning, then that is a design flaw almost akin to a hardware deficiency; the bulk of buyers will not hear them in the intended manner.

I mean, maybe after some experimentation one might find the Kii Three actually sounds best when mounted on the sidewalls at a 27.3-degree diagonal angle, but if the manufacturer doesn't tell you that, it doesn't matter much.

Not saying what you're saying about special positioning is untrue, and I haven't seen the dipole studies you're referring to. But it strikes me as very odd that manufacturers wouldn't be more forthright about positioning, given almost every other high-end speaker I know of with unique placement requirements tells you so. Earlier in this thread we talked about HSU speakers requiring special positioning with extreme toe-in, and they tell you that in the manual. The Dutch and Dutch 8C, unlike most speakers, are meant to be coupled with the front wall and therefore placed close to it, and they also tell you that in the manual.

 

[Cosmik]

I can't understand his dislike of narrow cabinets. Eg. baffle loss/step is a very simple and easily handled phenomenom well know for ages

I don't quite see that. He is saying that the on-axis and off-axis sounds are very different, and your ears pick up both and are unable to reconcile them. There is nothing you can do with EQ to fix that as it is baked into the physical dimensions of the box and drivers.

You can do some baffle step compensation, but it is a compromise and not a correction.

 

[Krunok]

I've seen the above mentioned elsewhere, but in section 5.3 of Toole's book the early reflections curve is actually characterized as an average of:

• Floor bounce: average of 20 °, 30 °, 40 ° down
• Ceiling bounce: average of 40 °, 50 °, 60 ° up
• Side wall bounces: average of ± 40 °, ± 50 °, ± 60 °, ± 70 °, ± 80 ° horizontal
• Front wall bounce: average of 0 °, ± 10 °, ± 20 °, ± 30 ° horizontal
• Rear wall bounces: average of 180 °, ± 90 ° horizontal

That includes a lot more vertical data. So I'm wondering if I'm missing something - are there two standards for the curve?

I think this what you stated is what is measured for the total sound power.