Michael Børresen explains himself?

[Mart68]

I don't recall KEF ever straying off the path?

 

[Arindal]

I think KEF nowadays strive towards neutral afaik ?

I am not familiar with all of their current models, but have heard quite a few, and seen frequency response graphs hinting to the direction they prefer a pronounced downward-sloped response both on-axis, but more so in the room. Q11 Meta was discussed recently as an example, while some legacy models like Reference 207/2 were pretty linear, my conclusion would be they have changed their house-sound curve, around 2010 maybe?

B&W not so much they clearly choose the batman curve

That has not always been the case. Matrix 801 or some legacy DM600S1/S2 series loudspeakers were pretty linear. Maybe around the year 2000 something changed.

By ‘influential’ do you mean the manufacturers who have heavily marketed their designs for decades?

No, any company selling a considerable number of units, no matter through which marketing methods. I don't differentiate between those relying their marketing on technical specs (like ´1,000 Watts!´, ´12-45,000Hz!´ or ´flattest frequency response´) or those promoting their products through other claims, may be it be exotic materials, technical solutions or whatever. It is all legitimate, and in the end of the day people are buying the speaker they subjectively like the most.

I am just observing that almost no successful high end speakers nowadays are marketed through purely technical specifications or measurements, so I would draw the conclusion not much has changed over the last 35 years when this type of marketing widely went out of fashion.

 

[Heinrich]

tl;dr

Best I can do is to remind us of our own gaps in knowledge and the lacking computation of which.

He falls for the term "frequency" and says that all frequencies have to have the same time -- somehow he's telling something.

The concept of frequency is a mathematical artifact, it is not true. Frequency eats time entirely, you have either time or frequency, never both. Not kidding about time this time.

Now knowing, do you still lough at him?

 

[FrantzM]

tl;dr

Best I can do is to remind us of our own gaps in knowledge and the lacking computation of which.

He falls for the term "frequency" and says that all frequencies have to have the same time -- somehow he's telling something.

The concept of frequency is a mathematical artifact, it is not true. Frequency eats time entirely, you have either time or frequency, never both. Not kidding about time this time.

Now knowing, do you still lough at him?

Emphasis is mine..
Care to elaborate?
What is "not true" about frequency?

 

[singular]

This debate is debacle.

90% HEA loudspeakers are BS based on +/-5% tolerance.
10% gap is way out of what is ideal.
This industry works with random results by luck.

In series the only way is to measure all what have two contacts...
Or DIFY.

 

[kawauso]

I am not familiar with all of their current models, but have heard quite a few, and seen frequency response graphs hinting to the direction they prefer a pronounced downward-sloped response both on-axis, but more so in the room. Q11 Meta was discussed recently as an example, while some legacy models like Reference 207/2 were pretty linear, my conclusion would be they have changed their house-sound curve, around 2010 maybe?



That has not always been the case. Matrix 801 or some legacy DM600S1/S2 series loudspeakers were pretty linear. Maybe around the year 2000 something changed.



No, any company selling a considerable number of units, no matter through which marketing methods. I don't differentiate between those relying their marketing on technical specs (like ´1,000 Watts!´, ´12-45,000Hz!´ or ´flattest frequency response´) or those promoting their products through other claims, may be it be exotic materials, technical solutions or whatever. It is all legitimate, and in the end of the day people are buying the speaker they subjectively like the most.

I am just observing that almost no successful high end speakers nowadays are marketed through purely technical specifications or measurements, so I would draw the conclusion not much has changed over the last 35 years when this type of marketing widely went out of fashion.

The tonality of the R5 and the Reference 203 seems similar, and their slopes also look alike.
The treble on the 203/2 and R300 had a slightly shallower slope and sounded a bit sharp to my ears, so it seems that some changes were made around that generation.

 

[Heinrich]

Emphasis is mine..
Care to elaborate?
What is "not true" about frequency?

The concept of frequency: count per time interval. That's naive in some sense. It implicates regularity and by that states perfect--rule based, knowledge of all the past and future. Basically it exists by definition, is a quite useful tool in science, but may become misleading when applied to streaming "signals", the music and the hearing. (In physics the meaning of frequency as a measured property is limited by Heisenberg, for instance.)

Edit: I didn't want to educate anyone, and the talking in the video is amusing at times, but we shouldn't redicule someone because of lacking knowledge to easily either.

 

[Arindal]

The tonality of the R5 and the Reference 203 seems similar, and their slopes also look alike.

I meant different generations and different series, as the products which made it obvious to me that their target curves were deviating from one another although being released just a few years apart from each other, were Reference 207/2 and Reference 3 or Reference 5. One being released before 2010, the other 2 shortly after. You are referring to 203 (previous generation, smaller model) and R5 (different series).

Don´t think the calculated in-room-response is very accurate in this case, as real-world results are heavily depending on room, listening distance and placement. Nevertheless there are examples of significantly downward-sloped responses which exemplarily support my claim:

I personally would assume the existence of a slope decreasing towards higher frequencies in SPL per -5dB or -7dB within in the localizable bands (taken bands in bass and upper treble into account, we are rather talking about -10dB) is indicative of a certain ´house curve´ or desired tonality ideal which I would not call linear.

Not saying that this constitutes bad speakers, but I don't understand why one ´house curve´ with mainly narrow-banded peaks and dips, in this case Børresen´s, alone is a reason for all sorts of disparagements and hostile reviews w/o listening tests, while other house curves which are more broad-banded in their attenuated or boosted frequency bands, which I would see as an indication of deviating stronger from an ideal of linearity, file under ´perfectly linear´.

 

[Keith_W]

Emphasis is mine..
Care to elaborate?
What is "not true" about frequency?

In signal processing, we think of the time domain and frequency domain differently. The frequency response is the Fourier transform of the impulse response, and the impulse response is the inverse Fourier of the frequency response. It's kind of abstract and it took me ages to wrap my head around it. Being a loudspeaker designer, I doubt if Borreson knows much about signal processing. And I don't know what point he is making. It may be another pseudoscientific point where he mixes truth and fiction to bamboozle his audience. We saw an example of that earlier in the thread.

 

[kawauso]

I meant different generations and different series, as the products which made it obvious to me that their target curves were deviating from one another although being released just a few years apart from each other, were Reference 207/2 and Reference 3 or Reference 5. One being released before 2010, the other 2 shortly after. You are referring to 203 (previous generation, smaller model) and R5 (different series).

Don´t think the calculated in-room-response is very accurate in this case, as real-world results are heavily depending on room, listening distance and placement. Nevertheless there are examples of significantly downward-sloped responses which exemplarily support my claim:
View attachment 494601

View attachment 494602

I personally would assume the existence of a slope decreasing towards higher frequencies in SPL per -5dB or -7dB within in the localizable bands (taken bands in bass and upper treble into account, we are rather talking about -10dB) is indicative of a certain ´house curve´ or desired tonality ideal which I would not call linear.

Not saying that this constitutes bad speakers, but I don't understand why one ´house curve´ with mainly narrow-banded peaks and dips, in this case Børresen´s, alone is a reason for all sorts of disparagements and hostile reviews w/o listening tests, while other house curves which are more broad-banded in their attenuated or boosted frequency bands, which I would see as an indication of deviating stronger from an ideal of linearity, file under ´perfectly linear´.

It’s true that the slope varies by generation and model, but saying KEF doesn’t keep their house curve consistent feels like an overstatement. The smoothness and dispersion characteristics remain quite similar.

I’m also not sure why the Q11 Meta example was brought up. My understanding is that the deeper downward slope comes from tuning the Uni-Q for its dispersion behavior and high-frequency range within the Q-series design, not from a fundamentally different “target.”

If two speakers have similar predicted in-room responses and similar directivity behavior, wouldn’t they end up with comparable FR in the same room?

 

[singular]

Being a loudspeaker designer, I doubt if Borreson knows much about signal processing. And I don't know what point he is making.

What point you are refering to?
Is MB in thread?

 

[sigbergaudio]

I meant different generations and different series, as the products which made it obvious to me that their target curves were deviating from one another although being released just a few years apart from each other, were Reference 207/2 and Reference 3 or Reference 5. One being released before 2010, the other 2 shortly after. You are referring to 203 (previous generation, smaller model) and R5 (different series).

Don´t think the calculated in-room-response is very accurate in this case, as real-world results are heavily depending on room, listening distance and placement. Nevertheless there are examples of significantly downward-sloped responses which exemplarily support my claim:
View attachment 494601

View attachment 494602

I personally would assume the existence of a slope decreasing towards higher frequencies in SPL per -5dB or -7dB within in the localizable bands (taken bands in bass and upper treble into account, we are rather talking about -10dB) is indicative of a certain ´house curve´ or desired tonality ideal which I would not call linear.

Not saying that this constitutes bad speakers, but I don't understand why one ´house curve´ with mainly narrow-banded peaks and dips, in this case Børresen´s, alone is a reason for all sorts of disparagements and hostile reviews w/o listening tests, while other house curves which are more broad-banded in their attenuated or boosted frequency bands, which I would see as an indication of deviating stronger from an ideal of linearity, file under ´perfectly linear´.

Kef Reference 5 seem to have around 7-8dB slope from 50hz to 10khz in the estimated response. This is likely to sound pretty linear (which is also my experience after hearing it).

 

[Geert]

Kef Reference 5 seem to have around 7-8dB slope from 50hz to 10khz in the estimated response. This is likely to sound pretty linear (which is also my experience after hearing it).

That's to be expected if you listen to KEF's chief engineer. No application of dubious or controversial theories there.

 

[ckernel8]

For home speakers, for that budget, you can achieve any level of performance or tonality you want, top-end room treatment, and a very comfortable couch to listen from.

...and the house too.

 

[MAB]

It sounds like the NS-18 has a strange presentation, but you actually seem to enjoy it.

I actually don't like the way they sound. They are in our spare bedroom and play classical music in the background for my mom and her dog. I joke that the dog prefers classic rock. But seriously, my mom has very little hearing left but likes some background music, she likes the look and operation of the old Yamaha integrated amp driving them, and they look the part for her age and the decor of the room. And her dog has some anxiety that music helps with.

Considering the price of Børresen speakers, the NS-18 feels like a surprisingly charming bargain.
I’m not sure how much you paid for the NS-18, but was it around ten dollars?

I got them from a local person to person sale for very cheap. They are incredibly high quality, which I discuss in the review thread that I linked above. The walnut plywood is top-notch. The wiring and interior details are completely sorted out. The drivers are fantastic Yamaha pieces right down to the coated Styrofoam woofer. I have nice speakers to listen to, these NS-18 are for a different purpose.

 

[Arindal]

but saying KEF doesn’t keep their house curve consistent feels like an overstatement.

As mentioned, I referred to particular Reference series models and I have listened to them side by side. In my understanding the tonality has changed quite significantly around 2010. And even if current models are not all the same, they in my ears more or less show a tendency towards the downward-sloped house curve.

The smoothness and dispersion characteristics remain quite similar.

I don't really know which result of a listening tests exactly defines ´smoothness´. Interestingly, such term is associated with the Børresens, which do sound overly ´smooth´ and ´silky´ in my understanding, to a degree I would not call them neutral. Same with the result of the dispersion characteristics of increasing directivity index towards higher frequencies. It contributes to dull reverb, bass/lower-midrange-heavy overall tonality. Some people might like the former, some the latter, I wouldn't call any variant ´neutral´ or ´natural´. Just fail to understand why one is a reason to disparage the designer, and the other one receives nothing but praise from technical/measurement perspective.

Reference 5 seem to have around 7-8dB slope from 50hz to 10khz in the estimated response. This is likely to sound pretty linear (which is also my experience after hearing it).

A bit cautious when it comes to bass response of a calculated in-room curve based on fixed parameters into account, as this is always far from an existing room with its room modes and boundary reflections.

But even if we are solely looking at localizable frequency bands (300-8,000Hz), the in-room response is sloped downwards in the -5dB region. In existing rooms with more of treble absorption and diffuse reverb dominant, this can be even more. Have never listened to an example in this ballpark which I would call linear, the lack of brillance and treble is quite obvious.

 

[kawauso]

As mentioned, I referred to particular Reference series models and I have listened to them side by side. In my understanding the tonality has changed quite significantly around 2010. And even if current models are not all the same, they in my ears more or less show a tendency towards the downward-sloped house curve.



I don't really know which result of a listening tests exactly defines ´smoothness´. Interestingly, such term is associated with the Børresens, which do sound overly ´smooth´ and ´silky´ in my understanding, to a degree I would not call them neutral. Same with the result of the dispersion characteristics of increasing directivity index towards higher frequencies. It contributes to dull reverb, bass/lower-midrange-heavy overall tonality. Some people might like the former, some the latter, I wouldn't call any variant ´neutral´ or ´natural´. Just fail to understand why one is a reason to disparage the designer, and the other one receives nothing but praise from technical/measurement perspective.



A bit cautious when it comes to bass response of a calculated in-room curve based on fixed parameters into account, as this is always far from an existing room with its room modes and boundary reflections.

But even if we are solely looking at localizable frequency bands (300-8,000Hz), the in-room response is sloped downwards in the -5dB region. In existing rooms with more of treble absorption and diffuse reverb dominant, this can be even more. Have never listened to an example in this ballpark which I would call linear, the lack of brillance and treble is quite obvious.

Smoothness refers to Tonality in the Spinorama context.
It has nothing to do with the “smoothness” that Børresen describes as a sonic texture.

KEF does not intentionally introduce odd Q-shaped features in their response.
Their approach to dispersion is consistent across generations and models:
they aim for smooth, continuous off-axis behavior.

If someone feels that KEF’s downward-tilted tonality is too warm or not completely neutral,
it can be easily corrected with EQ.

In contrast, Børresen exhibits discontinuities in directivity caused by resonances.
These discontinuities make EQ correction difficult, even for something as simple as a slight tilt.
This is not an issue with KEF-type speakers, where the behavior is inherently well-controlled.

 

[Arindal]

Smoothness refers to Tonality in the Spinorama context....they aim for smooth, continuous off-axis behavior.

So you define ´Spinorama smoothness´ as a technical description for continously increasing directivity index towards higher frequencies? That's a technical, or rather a geometrical definition when looking at the graph - when listening to music on a speaker in an existing room, how do I exactly hear it, how does it translate to subjective impression? And why should it be superior to either linear indirect sound (as a result from constant directivity) or just narrow-banded/mild/random deviations in the indirect soundfield?

Fully understood that ´smoothness´ is defined in several contradictive ways, and the ´Børresen style smoothness´ is rather a subjective description than a technical definition.

If someone feels that KEF’s downward-tilted tonality is too warm or not completely neutral,
it can be easily corrected with EQ.

That is not only counterintuitive, it is also destined to fail on a practical level (and I really tried exactly that at home). If you have linear direct sound response and significantly decreasing level of indirect sound towards higher frequencies, an EQ cannot correct for this increasing difference between the two. In contrary, trying to correct a broad-banded problem which exists solely in the indirect soundfield, by EQ, like lack of treble and brilliance level, you end up with boosting and subjectively emphasizing this band in the direct sound. Which we hopefully all agree on will lead to kinked tonality and would be far from neutral.

Børresen exhibits discontinuities in directivity caused by resonances.

I don't mean to defend a manufacturer whose products I subjectively don't like and find questionable, but I don't see dramatic discontinuity of directivity caused by resonances. With the model that has been measured by Erin (X3), there is seemingly a sharp on-axis peak at 2.2K, but it is very narrow, corresponds with higher directivity index and shows zero signs of resonances. But rather lobing or cancellation issues of one or several stiff diaphragms being driven to higher frequency bands than they could handle without directivity problems, so under different angles the problem almost disappears and chances are pretty good that you can EQ exactly that.

 

[sigbergaudio]

As mentioned, I referred to particular Reference series models and I have listened to them side by side. In my understanding the tonality has changed quite significantly around 2010. And even if current models are not all the same, they in my ears more or less show a tendency towards the downward-sloped house curve.



I don't really know which result of a listening tests exactly defines ´smoothness´. Interestingly, such term is associated with the Børresens, which do sound overly ´smooth´ and ´silky´ in my understanding, to a degree I would not call them neutral. Same with the result of the dispersion characteristics of increasing directivity index towards higher frequencies. It contributes to dull reverb, bass/lower-midrange-heavy overall tonality. Some people might like the former, some the latter, I wouldn't call any variant ´neutral´ or ´natural´. Just fail to understand why one is a reason to disparage the designer, and the other one receives nothing but praise from technical/measurement perspective.



A bit cautious when it comes to bass response of a calculated in-room curve based on fixed parameters into account, as this is always far from an existing room with its room modes and boundary reflections.

But even if we are solely looking at localizable frequency bands (300-8,000Hz), the in-room response is sloped downwards in the -5dB region. In existing rooms with more of treble absorption and diffuse reverb dominant, this can be even more. Have never listened to an example in this ballpark which I would call linear, the lack of brillance and treble is quite obvious.

As I think have been discussed before, I'm not sure how accurate those in-room responses are. Looking at my own designs, the estimates show a similar slope (admittedly more shallow), while in reality they're often quite flat in-room from 3-4-500hz and up to 7-10khz (where the tweeter naturally starts to slope due to listening distance).

I have never measured Kef reference in-room, but I actually have measured the LS60, so perhaps I can try to find that later and compare with the estimated response for fun.

 

[kawauso]

As I think have been discussed before, I'm not sure how accurate those in-room responses are. Looking at my own designs, the estimates show a similar slope (admittedly more shallow), while in reality they're often quite flat in-room from 3-4-500hz and up to 7-10khz (where the tweeter naturally starts to slope due to listening distance).

I have never measured Kef reference in-room, but I actually have measured the LS60, so perhaps I can try to find that later and compare with the estimated response for fun.

I compared your company’s speakers with KEF, and they seem quite similar.
Sigberg Audio looks like the kind of speaker I would personally enjoy.
The reduced bass appears intentional, probably because they are designed to be used with subwoofers.

From around 1000 Hz and below, the dispersion pattern appears sharply cardioid.
Is there any active DSP correction involved in achieving that dispersion behavior?
It also looks like they would deliver bass with the kind of punch you normally get from larger speakers.