KEF/Coaxial 2 Ways vs 3 way point source qualities

[thewas]

Here I don't see any hint on how crucial a mere +/-1dB in directivity is, let alone a 'scientific' argument why it would be so.

That was my response to your question about "people listening to rooms".

I stepped in because, well, I think some, and you also, exaggerate things a bit on a regular basis.

No, you rather tend to misinterpret my and other people comments on a regular basis. I never for example said those differences are large but audible which is a very different thing.

Back on topic, you say that difference between LS50 and R3 are due to the directivity. Not really with me--for sure with you!
I say, differences between LS50 and R3 are due to the output capabilities, especially when playing Daft Punk. The other way round.

Same here, I never mentioned Daft Punk in any of my statements but said the difference with not too high SPL and/or too bass heavy music which is exactly the opposite.

In somehow denying my argument implicitely you cross one of my mantras that I reiterate all the time. An imperfect speaker adds drama, which some people simply like. I don't argue against that, it's just a provocative suspicion. I leave people their preferences (but, it scrutinizes the Harman diliberately *g*).

Both the R3(M) and LS50(M) have clear limitations and imperfections, a bit different ones though and both still are absolutely fine, even more for their price classes.

What we without any doubt agree on is, that a comparison R3 versus LS50 won't decide, if an added bass it disliked because of its separation to the mid-speaker. What do you think?

I didn't go into the subjective preference as this depends on many different as well as also individual factors, just gave some reasons why they sound different and assumptions why one could prefer in a specific scenario one to another.

 

[Heinrich]

No, you rather tend to misinterpret my and other people comments on a regular basis. I never for example said those differences are large but audible which is a very different thing.

You said this exactly in the referring post #67:

Yes, the drawback of the R3 isn't .. but rather its ... which leads to less continuous directivity, ...
In my experience this is what makes the R3 sometimes not being preferred at a direct comparison.

So, you say that for some, especially yourself, parameter "directivity" of the R3 is comparatively worse so much, that it affects preference. So it is decidedly relevant, whilst only +/-1dB. The latter is less than the typical on-axis ripple even for studio monitors. The 'misinterpretation' you accuse me of isn't there.

You tried to refer the above and other argumentation to Dr. Toole's contribution in post #1953 (!) in some other thread ("Genelec 8030C ..."). With all due respect I couldn't find anything supporting dogma. Directivity is a factor, admitted, but pulling a +/-1dB ripple into a serious discussion, appears as what I named it, a severe loss of proportion. What I didn't mention here, is the differences in studio equipment that copy over to the individual recordings, let alone the sound engineers daily mood.

I offered you an open hand, which you discarded. My mood went sour with this conversation, and I wrap it up.

 

[Jazigo]

Admittedly, I don't know enough to actually contribute in a meaningful way, but a subwoofer is localize-able. It seems that 100hz an below is simply not localize-able, but everything is above that is.

I find this topic very interesting, so it would be nice to help each other on the way to acquire more knowledge about this.

 

[aarons915]

You get your applause, fair enough again. I question the plausibility of your claim, me coming from a lot of years in the hobby, and owning and designing a lot of very different speakers.

First step first, you say: "The easiest comparison is to A/B an LS50 or Q1/3 to a Q11 or R11 ..."

When you did that, was it blindfolded? Not that it makes a difference. As I elaborated on already, those speakers are vastly different to begin with. Not only the timbre, but the spl cpabilities alone would make them easily distinguishable even blindfolded. Any bias is so as easily copied over to the "test". This combines with the test done only once by one person in one environment. Starting with the latter, we do not have the slightest idea what the surrounding was, what your preferences are, e/g in timbre, the signal and so on ad nauseam. In short, your experience is stuck to you. It's an individual memory you actually cannot share with others. There is no inter-personal communication possible regarding your test.

Then you formulate a claim based on that experience, that in its main argument contradicts well established knowledge. (Perhaps take a read on vertical localisation.)

Back to plausibility, we might consider other, non-coaxial loaded multi-way designs. The pretty old standard. Anyone to "hear"--means: not see, the distance between the individual drivers as different locations, in the vertical?

How could that be: the bass distorts, which in essence is radiating higher frequencies than intended. Maybe that may be a root cause for localisation. We might as well out-rule this possibility as the R3 is pretty much clean, virtually distortion-free in the suspected registers. Any other, I don't know, your turn.

In audio we are dealing with the so called psycho-acoustics, a nasty word, admitted. We have to consider at least two facets, that reflect very complicated topics. The human perception as a whole, and a vastly complicated soundfield.

So your claim may be valid, but to just tell lays the burden on others, while you could and should be the one to educate others on the very topic. Please, go ahead.

Again, I'm not arguing what should or shouldn't be, just saying what I've heard and to answer your questions I've noticed this both blind and sighted and in different rooms with different speakers involved. I think we all get that the speakers are tonally different with different capabilities, if you go back you'll note I wasn't discussing any of that, only that splitting the midrange between a woofer and midrange doesn't sound as cohesive as the point source capabilities of the 2 way coax drivers. I believe I've mentioned that I barely heard this effect in the Reference Ones which is what leads me to believe it's related to the crossover frequency, the original reference were crossed over lower if I recall correctly. I believe if KEF were to design their speakers with a crossover closer to 200Hz for the midrange this wouldn't be an issue at all.

Side note since you've mentioned a few times about directivity of 1-2db not being very audible, the R3 are one example where I noticed a problem with harshness and I tracked it down to the area between 1-3k. Most people looking at the spin wouldn't notice an issue but I ended up having to smooth out the early reflections, which puts a slight dip in the listening window response in order to smooth them out and sound more natural, you can see the directivity is increasing and then comes back down in that area making for a stronger off-axis response.

I love theory too but we don't have a ton of research on this stuff so I think it would do you well to do some actual listening tests on this stuff. I've had my main to sub crossover as high as 150Hz because I couldn't localize my subs when I actually tested it out, if I listened to the consensus most people would say I'm crazy and anything over 80Hz is going to be no good.

 

[Heinrich]

... noticed this both blind and sighted and in different rooms with different speakers involved. I think we all get that the speakers are tonally different with different capabilities, if you go back you'll note I wasn't discussing any of that, ...

Exactly. It was me to introduce the identifiability due to the differnt sound profiles. That renders a blindfolding useless. You might have trimmed at least the on/off-axis responses to a similar profile in order to concentrate the audition more on what you were after ... . Obviously you didn't catch my argument.

... only that splitting the midrange between a woofer and midrange doesn't sound as cohesive as the point source capabilities of the 2 way coax drivers. ...

I feel that your personal observations under non-focused aka unconolled conditions are not easily generalizable.

I love theory too but we don't have a ton of research on this stuff ...

We have. And that's the point, in my book at least Your question targets well established knowledge that reaches from plain physics over physiology to the simple fact, that regular non-coaxial multi-way speakers don't show the effect you personally experience. As you said in your post #1, nobody else seems to complain about the downfalls you tempt to expose.

I wanted to bring something up that most people never talk about ...

Now people are talking, right? Coming back to research. Do you implicitely want science to clarify the issue--in case, what hinders you to just do it? Find an agreeable methodology, we might help with that.

 

[aarons915]

Exactly. It was me to introduce the identifiability due to the differnt sound profiles. That renders a blindfolding useless. You might have trimmed at least the on/off-axis responses to a similar profile in order to concentrate the audition more on what you were after ... . Obviously you didn't catch my argument.


I feel that your personal observations under non-focused aka unconolled conditions are not easily generalizable.


We have. And that's the point, in my book at least Your question targets well established knowledge that reaches from plain physics over physiology to the simple fact, that regular non-coaxial multi-way speakers don't show the effect you personally experience. As you said in your post #1, nobody else seems to complain about the downfalls you tempt to expose.


Now people are talking, right? Coming back to research. Do you implicitely want science to clarify the issue--in case, what hinders you to just do it? Find an agreeable methodology, we might help with that.

And is there any feasible theory as to why different tonality would be responsible for what I'm talking about? It really makes no sense and by the way KEF speakers mostly sound similar anyway, the SPL capabilities only matter if I'm pushing the 2 ways into distortion which I'm not so again that theory makes no sense...

A few people have noticed the same thing as me, maybe you should actually try to conduct a similar test instead of coming up with random theories that aren't even applicable? I also noted that it's not something you notice unless you directly compare as I and a few others have said. I literally had the R3 for a year and didn't notice it until I tried another speaker and directly compared them blind, so it's not a shocker that most people don't talk about this.

Regular multi way speakers absolutely have these downfalls, people don't talk about them because it is considered normal and what they're used to. If you compared a typical multi way to an LS50 or other 2 way coaxial, there would be an obvious difference where the sound is coming from...you should really do this before repeating your theories for the 5th time...

 

[Heinrich]

And is there any feasible theory as to why different tonality would be responsible for what I'm talking about?

Yes, humans are prone to prejudice, bias. The speakers sound very different, identifiable different. When you audition them for differences in terms of driver distance, you would know which you listen to even if blindfolded. That way the bias from the looks, separated drivers namely, is spread over to the audition.

I told you the third/forth time? You didn‘t pick up the argument at any occasion, but pretend it would be incomprehensable. I already told you how to mitigate the problem, as to explicate it practically, no reaction, no respect. Do you make fun of me?

What else to say? At least it would me work way way harder than you. So, nothing me thinks, you win.

 

[aarons915]

Yes, humans are prone to prejudice, bias. The speakers sound very different, identifiable different. When you audition them for differences in terms of driver distance, you would know which you listen to even if blindfolded. That way the bias from the looks, separated drivers namely, is spread over to the audition.

I told you the third/forth time? You didn‘t pick up the argument at any occasion, but pretend it would be incomprehensable. I already told you how to mitigate the problem, as to explicate it practically, no reaction, no respect. Do you make fun of me?

What else to say? At least it would me work way way harder than you. So, nothing me thinks, you win.

I understand what you're saying agree with you, I get bias exists but that really isn't relevant to what we're discussing. I'm not even saying one is better than the other even though I personally prefer the point source, when I was listening to the Q6 they sounded more like a "wall of sound" and I could definitely see some people preferring that. When I listen for this effect between speakers, I'm not listening for preference, I'm solely focused on identifying the source of the sound and simply noticing a difference. You're going off on all of these tangents that aren't really relevant to what I'm talking about and no you didn't tell me how to mitigate the effect, even if listening to them nearfield fixed the issue (it doesn't), that's not how I use mine.

Another piece of evidence I don't think I mentioned is the previous Q series towers were 2.5 way and they retained this point source characteristic, so to me that shows that even if some of the upper bass/midrange is sent to the woofer as long as the UniQ plays low enough it will retain this characteristic. Note I have said nothing about the sound quality of any of these comparisons I've done as I'm only focusing on this issue with the comments I'm making.

Another easy way of testing this would be to make a quasi 3 way speaker with one of these 2 way Coaxial speakers and gradually increase the crossover between them, I suspect as you go up to 200,300, 400, 500, etc, you will notice a shift in vertical localization or a blurring of this image as the higher frequencies in the sub become more localizable.

 

[Jazigo]

Ok, so I have read through all the posts in this thread and I have tried to pieced things together:

• Horizontal localization something we can easily do, even at low frequencies.
• Vertical localization is easy at higher frequencies, but very hard at lower frequencies.

This means it is unlikely of us actually hear the vertical offset of the woofer on the R3.

Could the lobe from the UniQ driver be affected by lobe of the woofer in a significant way?

Also, the woofer and the UniQ drivers have different vertical distance to the top and the bottom of the case, could that introduce a noticeable difference?

 

[aarons915]

Ok, so I have read through all the posts in this thread and I have tried to pieced things together:

• Horizontal localization something we can easily do, even at low frequencies.
• Vertical localization is easy at higher frequencies, but very hard at lower frequencies.

This means it is unlikely of us actually hear the vertical offset of the woofer on the R3.

Could the lobe from the UniQ driver be affected by lobe of the woofer in a significant way?

Also, the woofer and the UniQ drivers have different vertical distance to the top and the bottom of the case, could that introduce a noticeable difference?

So I just did the test I described, I use Equalizer APO on windows so I can increase the low pass on the sub to whatever I want, my denons LFE is limited to 250Hz though but even at 250Hz you can clearly hear a shift of vocals from the LS50 to hearing them in the sub as well. I understand the research says that vertical localization only occurs in higher frequencies but this is why I say we shouldn't be just blindly accepting the research when it's easy to test this stuff out.

 

[Jazigo]

So I just did the test I described, I use Equalizer APO on windows so I can increase the low pass on the sub to whatever I want, my denons LFE is limited to 250Hz though but even at 250Hz you can clearly hear a shift of vocals from the LS50 to hearing them in the sub as well. I understand the research says that vertical localization only occurs in higher frequencies but this is why I say we shouldn't be just blindly accepting the research when it's easy to test this stuff out.

True! It can be a very viable way to challenge that.

Currently, I am reading up how a tweeter and woofer effect each other in a classical two-way speaker. Maybe that knowledge can help shape another viable way to explain this.

 

[Jazigo]

So I just did the test I described, I use Equalizer APO on windows so I can increase the low pass on the sub to whatever I want, my denons LFE is limited to 250Hz though but even at 250Hz you can clearly hear a shift of vocals from the LS50 to hearing them in the sub as well. I understand the research says that vertical localization only occurs in higher frequencies but this is why I say we shouldn't be just blindly accepting the research when it's easy to test this stuff out.

Wait a second, you where actually able to hear vocals on the subwoofer? Could you try to move the subwoofer up - say 30 centimeter from the floor? If you can hear a very notable difference from 0cm to 30cm, maybe you could try with 15 cm? I guess at one point, one can't hear the difference.

 

[aarons915]

True! It can be a very viable way to challenge that.

Currently, I am reading up how a tweeter and woofer effect each other in a classical two-way speaker. Maybe that knowledge can help shape another viable way to explain this.

I've heard typical 2 ways still have a pretty good point source presentation so it's my belief that you want to keep the fundamental frequencies, especially vocals, coming from a single source. The highs coming from the tweeter don't seem to be a distraction.

Wait a second, you where actually able to hear vocals on the subwoofer? Could you try to move the subwoofer up - say 30 centimeter from the floor? If you can hear a very notable difference from 0cm to 30cm, maybe you could try with 15 cm? I guess at one point, one can't hear the difference.

I heard vocals when the sub was playing up to 250Hz, I don't normally hear them. Most KEF 3 ways have the woofers playing up to 500Hz though so even higher frequencies.

I didn't try to elevate the subs but if I can hear a difference with a Q6 I'm sure it would be no different as the spacing is only like 6 inches between woofer and UniQ. Heinrich is right that it's not necessarly vertical localization that I'm hearing, it's more like a smearing or blending of vocals between sources which just sounds a bit different than everything coming from the UniQ driver, the image is still anchored to the center because of the highs.

 

[Jazigo]

I've heard typical 2 ways still have a pretty good point source presentation so it's my belief that you want to keep the fundamental frequencies, especially vocals, coming from a single source. The highs coming from the tweeter don't seem to be a distraction.



I heard vocals when the sub was playing up to 250Hz, I don't normally hear them. Most KEF 3 ways have the woofers playing up to 500Hz though so even higher frequencies.

I didn't try to elevate the subs but if I can hear a difference with a Q6 I'm sure it would be no different as the spacing is only like 6 inches between woofer and UniQ. Heinrich is right that it's not necessarly vertical localization that I'm hearing, it's more like a smearing or blending of vocals between sources which just sounds a bit different than everything coming from the UniQ driver, the image is still anchored to the center because of the highs.

I see that KEF R3 Meta, which were to ones I tested is crossed over at 420hz. That is a bit higher than I thought. On the Edifiers 1280 I use at work, I can clearly hear the difference, but those are used in near-field and are crossed over at 2.5 khz.

When I A/B tested the R3- and LS50 Meta, the distance from me to the speakers was about 2 and half meters. It was in a room that was acoustically treated with bass-traps and panels.

Maybe 420hz is actually high enough to hear the vertical offset of the woofer.

 

[aarons915]

I see that KEF R3 Meta, which were to ones I tested is crossed over at 420hz. That is a bit higher than I thought. On the Edifiers 1280 I use at work, I can clearly hear the difference, but those are used in near-field and are crossed over at 2.5 khz.

When I A/B tested the R3- and LS50 Meta, the distance from me to the speakers was about 2 and half meters. It was in a room that was acoustically treated with bass-traps and panels.

Maybe 420hz is actually high enough to hear the vertical offset of the woofer.

I think so, I've wondered if an LS50 meta driver crossed over in the 150-200Hz range would be the best of both worlds

 

[Ssss]

I think so, I've wondered if an LS50 meta driver crossed over in the 150-200Hz range would be the best of both worlds

I cross over my LS50Meta’s at 200hz & very satisfied with the sound, they are supported by a pair of 2M tall 4x15” bass arrays though.

 

[Heinrich]

In case some disbeliever is willing to take the effort testing the hopeless claim made, it should be worth it.

My two cents on that—isolate the insinuated effect, report.

Hypothesis: …, and so a difference is identified, when playing … music, speech, …

Build two speaker boxes 20x40sqcm baffle, one with a centered single driver, size 7“, the other with same type but two in close proximity, crossed over using LR type 12dB/oct. Check amplitude response for equivalence, off-axis included, room response included, document. Double check distortion, rattle, and foremost intermodulation, report.

Play speech music etc bandlimited as to avoid IM, HD to contaminate the outcome, evaluate blindfolded, if effect is identifiable in between the speakers. Redo on different occasions, ask other to participate

Such would be the bare minimum, before picking fear, uncertainty and doubt out of the blue attacking people‘s peace of mind like a flashing-down falcon. As a single person speaking against established engineering and the general experience.

 

[aarons915]

Such would be the bare minimum, before picking fear, uncertainty and doubt out of the blue attacking people‘s peace of mind like a flashing-down falcon. As a single person speaking against established engineering and the general experience.

I don't know if anyone is going to go through all that trouble but it's a good suggestion. And this is a bit extreme Heinrich lol...I don't think anyone feels like their peace of mind is attacked, it's just a friendly discussion asking if others have similar experiences.

 

[slaweks]

FWIW I replaced my LS50 Meta in the desktop system with modded d'Appolito design Pioneer SP-C22 speakers for the greater 'height' of the stereo image. The Meta's went into the home theater system downstairs. It looks like I want to look into the Genelec's if upgrading.

But keep in mind that Genelecs are hissing.

 

[Heinrich]

I don't know if anyone is going to go through all that trouble but it's a good suggestion.

Good suggestion, well, it's brilliant. What else would you ask for? People talking "similar experiences", sighted, uncontrolled, severely biased due to lack of understanding of the very basics, not to the point observations--or imaginations, mixed up with everything else possible.
Only to defend an ideal, that doesn't exsist: the "point source", which is a misnomer to begin with.

That's not a rant. I feel it pretty much describes what is happening here. Let alone the unavoidable side note on "directivity" with nearly every topic. I strongly advocate to return to rational waters. That is, state a claim, try to prove it wrong and introduce others to do the same. As you mentioned, there's a bit of an effort involved. If people aren't willing to take it, fair enough, but why then alll the talking?!

And this is a bit extreme Heinrich lol...I don't think anyone feels like their peace of mind is attacked, it's just a friendly discussion ...

It is not. Without going into analysis of social behaviour, it is about fear mongering, that the stereo wasn't sufficiant--due to what ever meniscule detail it will fail to deliver. So it began with cables, cabinet materials, the exaggeration of directivity issues, linearity issues, HD and everything extremist, whilst in parts measurement driven, that poisioned the social discourse on stereo.