• 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!

Properties of speakers that creates a large and precise soundstage

Bjorn

Major Contributor
Audio Company
Forum Donor
Joined
Feb 22, 2017
Messages
1,313
Likes
2,602
Location
Norway
Broadband constant directivity is far more important in regards to these attributes. The directivity should be uniform down to at least 500 Hz, preferably both horizontally and vertically. Remaining the directivity very high in frequency is less important, though wide dispersion in the highs can attribute to better 3D experience. But something like the use of uBAACH will yield better 3D than anything else IMO.

Besides broadband constant directivity, a vertical direcitivty that minimizes floor and ceiling reflections is a great advantage.
Of course, one can treat the surfaces acoustically and get closer too, but it'a always and advantage to start with a good speaker design.

In regards to the coaxial design. I have tested many coxial drivers over the years. First of all, none of them have been truly time aligned and which leads to both comb filtering and lobing unless one adds delay. Therefore I would never recommend a coxial speaker with a passive crossover between the two drivers. And comb filtering and frequency deviations will still happen to most of them because the two will interefere with each other. Secondly, I have found that almost all coxial designs suffer from a bit high intermodulation distortion. And apart from one, I found all lacking in openness, detail and clarity. This makes them sound more like "good hifi" vs being able to portrait an illusion of real instrument and vocals.

So the design isn't without compromises, though the vertical directivity is obviously better than a speaker with a crossover in the presence area combined with spacing between the drivers. The latter is not really good either.
 

Vacceo

Major Contributor
Joined
Mar 9, 2022
Messages
2,682
Likes
2,833
Isn't that narrowing on the vertical axis while keeping the constant directivity in the horizontal what Perlisten aims to do?

It is curious how similar dispersion patterns of Perlisten and KEF look on the horizontal (not so much in the vertical) while being fairly different arrangements.

Still, DSP's and active systems can correct all those inherent problems, which seems to be precisely how Genelec monitors "do their magic".
 

kongkong

Member
Joined
Jan 13, 2023
Messages
49
Likes
75
Isn't that narrowing on the vertical axis while keeping the constant directivity in the horizontal what Perlisten aims to do?

It is curious how similar dispersion patterns of Perlisten and KEF look on the horizontal (not so much in the vertical) while being fairly different arrangements.

Still, DSP's and active systems can correct all those inherent problems, which seems to be precisely how Genelec monitors "do their magic".
Perlisten and Genelec the Ones series are kind of psuedo-coaxial speaker.
There are lots of interesting information.
If you want to know, recommend finding research papers or diy maker's loudspeaker.


Like above that.

:)
I also like unique engineering to make better directivity.


Here is The one's coaxial paper
 
Last edited:

DSJR

Major Contributor
Joined
Jan 27, 2020
Messages
3,416
Likes
4,573
Location
Suffolk Coastal, UK
Hi everyone, it has been a minute.
I am getting rid of my Genelecs 8341 to free up some cash, and I am thinking of replacing these either with a non-coaxial design such as Genelecs 8040b, or another coaxial design such as Kef's R3s.
I enjoy a lot the 3d sound staging that the 8341 can depict, which I remember also experiencing with my LS50s in the past. So I have been assuming that this is an attribute in which coaxial design inherently excels, but is this true?
I listen from a distance of about 2m.

Thanks!
View attachment 326221
Having in the past sold some much loved gear I now bitterly regret doing looking back (and can never again aspire to for financial and space reasons), is there any possible way you can find the needed revenue from other means and keep the speakers? Those monitors are pretty much at or very near the top of the tree in what they do and I honestly feel you'll regret selling them later on.

Good luck with any decision you make :)
 

thewas

Master Contributor
Forum Donor
Joined
Jan 15, 2020
Messages
6,904
Likes
16,935
In regards to the coaxial design. I have tested many coxial drivers over the years. First of all, none of them have been truly time aligned and which leads to both comb filtering and lobing unless one adds delay. Therefore I would never recommend a coxial speaker with a passive crossover between the two drivers.
Can you please show such comb filtering and lobing on some current passive KEF loudspeaker measurements where the acoustic centre of the tweeter is very close to the one of the mid driver? By the way this is often a bigger issue on many passive non-coaxial loudspeakers, so not really an argument against coaxial drivers.

Secondly, I have found that almost all coxial designs suffer from a bit high intermodulation distortion.
Don't agree there on good 3-way (mid-woofer) coaxial drivers, Erin does in the past months multitone distortion measurements and some newer Genelec and KEF 3-way coaxial loudspeakers show excellent results.
 

Bjorn

Major Contributor
Audio Company
Forum Donor
Joined
Feb 22, 2017
Messages
1,313
Likes
2,602
Location
Norway
Can you please show such comb filtering and lobing on some current passive KEF loudspeaker measurements where the acoustic centre of the tweeter is very close to the one of the mid driver? By the way this is often a bigger issue on many passive non-coaxial loudspeakers, so not really an argument against coaxial drivers.


Don't agree there on good 3-way (mid-woofer) coaxial drivers, Erin does in the past months multitone distortion measurements and some newer Genelec and KEF 3-way coaxial loudspeakers show excellent results.
I haven't tested the KEF coaxial driver. I have only measured a few KEF speakers. Here's the latest measurement I did of a pair of KEF Q350 with a subwoofer (left and right channel). The measurement is in the listening position though. 1/12 Oct. smoothing. You can see some obvious issues in the 1.5 kHz-2 kHz area. I measured the same dip and peak at several positions. I'll do a near field measurement when I got time. It's a setup at my second job place.

KEF Q350.jpg



I have never measured IMD of neither KEF or Genelec, but I have never found them to sound particular open, detailed or life/illusion like. There other speaker designs that are far better in that area IMO. Why, it's difficult to know the exact reason for. Whether it's some kind of distortion in the driver, interference or also related to electronics, crossover and type of correction that has been done. To me, these speaker designs sounds like decent hifi.

One should definetly go with the active version as previously mentioned. Even a small difference in time from the drivers will lead to lobing in the frequency area.
 

thewas

Master Contributor
Forum Donor
Joined
Jan 15, 2020
Messages
6,904
Likes
16,935
I haven't tested the KEF coaxial driver. I have only measured a few KEF speakers. Here's the latest measurement I did of a pair of KEF Q350 with a subwoofer (left and right channel). The measurement is in the listening position though. 1/12 Oct. smoothing. You can see some obvious issues in the 1.5 kHz-2 kHz area. I measured the same dip and peak at several positions. I'll do a near field measurement when I got time. It's a setup at my second job place.
Such a single listening position room measurement can show anything like diffraction issues, to show supposed timing and lobing issues anechoic measurements are needed and all the ones we have know don't such.
Please have a look also at the white paper chapter 4.1 "The tweeter is placed at the acoustic centre of the midrange driver.", which is also shown with corresponding measurements.

have never measured IMD of neither KEF or Genelec, but I have never found them to sound particular open, detailed or life/illusion like. There other speaker designs that are far better in that area IMO. Why, it's difficult to know the exact reason for. Whether it's some kind of distortion in the driver, interference or also related to electronics, crossover and type of correction that has been done. To me, these speaker designs sounds like decent hifi.
Those personal subjective impressions don't bring anything to such a discussion.

One should definetly go with the active version as previously mentioned. Even a small difference in time from the drivers will lead to lobing in the frequency area.
For that to be true you must first show that there is a time difference. Also as you know for your favorised non coaxial loudspeakers you can only time align one single angle and all others like from table, floor and ceiling reflections have much larger time differences which is also seen at the formed vertical lobes.
 

thewas

Master Contributor
Forum Donor
Joined
Jan 15, 2020
Messages
6,904
Likes
16,935
Isn't that narrowing on the vertical axis while keeping the constant directivity in the horizontal what Perlisten aims to do?

It is curious how similar dispersion patterns of Perlisten and KEF look on the horizontal (not so much in the vertical) while being fairly different arrangements.

Still, DSP's and active systems can correct all those inherent problems, which seems to be precisely how Genelec monitors "do their magic".
I must say I really like also the new approach of Perlisten making a real D'Appolito by using an array of actually tweeter drivers into the midrange.

Here is superb and more extreme implementation of such an acoustic coxial loudspeaker with high vertical beaming
 
  • Like
Reactions: MAB

Matias

Master Contributor
Forum Donor
Joined
Jan 1, 2019
Messages
5,089
Likes
10,949
Location
São Paulo, Brazil
I have found that almost all coxial designs suffer from a bit high intermodulation distortion. And apart from one, I found all lacking in openness, detail and clarity. This makes them sound more like "good hifi" vs being able to portrait an illusion of real instrument and vocals.
Why is it that Doppler IM is always mentioned with coax drivers, but not with a woofer in a 2 way design? A woofer has for example a large 60 Hz cone displacement modulating a 1 kHz signal being played in the same driver, while a coax has for example a 300 Hz signal on the mid modulating the 2+ kHz signals of the tweeter. The woofer should be a lot more severe than the coax, yet no one mentions that. What am I missing?
 
Last edited:

Bjorn

Major Contributor
Audio Company
Forum Donor
Joined
Feb 22, 2017
Messages
1,313
Likes
2,602
Location
Norway
@thewas
My comment was about coaxial designs that I've measured. And I have measured difference of time arrival in all of these with the resultance of lobing. And when you have lobing, you also have combing FIY.

As mentioned, I haven't measured the coxial that KEF and Genelec uses. Only some speakers. I did measure the Q350 speakers at several positions FIY and the dip at 1.5 kHz was there in all of them. But I would need to meausure the speaker without room contribution one day, to know for sure if it's actually speaker related.
 

thewas

Master Contributor
Forum Donor
Joined
Jan 15, 2020
Messages
6,904
Likes
16,935
@thewas
My comment was about coaxial designs that I've measured. And I have measured difference of time arrival in all of these with the resultance of lobing. And when you have lobing, you also have combing FIY.

As mentioned, I haven't measured the coxial that KEF and Genelec uses. Only some speakers. I did measure the Q350 speakers at several positions FIY and the dip at 1.5 kHz was there in all of them. But I would need to meausure the speaker without room contribution one day, to know for sure if it's actually speaker related.
Well if the dip was in several positions it wasn't a lobing issue. ;) Also as said there exist already tons of anechoic measurements of the current KEF and Genelec coaxials which show that the two general disadvantages you mentioned are not an issue.

Here even for the entry 2-ways model Q350 where it can be seen there is no real lobing:

SPL%20Horizontal%20Contour%20Normalized.jpg

SPL%20Vertical%20Contour%20Normalized.jpg
 
OP
sigbergaudio

sigbergaudio

Major Contributor
Audio Company
Forum Donor
Joined
Aug 21, 2020
Messages
2,712
Likes
5,725
Location
Norway

HarmonicTHD

Major Contributor
Joined
Mar 18, 2022
Messages
3,326
Likes
4,837
Well if the dip was in several positions it wasn't a lobing issue. ;) Also as said there exist already tons of anechoic measurements of the current KEF and Genelec coaxials which show that the two general disadvantages you mentioned are not an issue.

Here even for the entry 2-ways model Q350 where it can be seen there is no real lobing:

SPL%20Horizontal%20Contour%20Normalized.jpg

SPL%20Vertical%20Contour%20Normalized.jpg
Yes.

See also here:

Thread 'KEF Q350 Speaker Review'
https://audiosciencereview.com/forum/index.php?threads/kef-q350-speaker-review.13484/

I agree it is not clear if the dip is due to the rather dated and „cost effective“ design of the Q350 speaker or the coax (non coax speakers also show this if not done properly). Especially as I don’t see it in the R3 and R3 Meta Klippels from Amir and Erin. So the cause and effect is not clearly proven for me, but admittedly I haven’t had time to look into the details.
 

Bjorn

Major Contributor
Audio Company
Forum Donor
Joined
Feb 22, 2017
Messages
1,313
Likes
2,602
Location
Norway
Why is it that Doppler IM is always mentioned with coax drivers, but not with a woofer in a 2 way design? A woofer has for example a large 60 Hz cone displacement modulating a 1 kHz signal being played in the same driver, while a coax has for example a 300 Hz signal on the mid modulating the 2+ kHz signals of the tweeter. The woofer should be a lot more severe than the coax, yet no one mentions that. What am I missing?
I never mentioned doppler shift. That's mainly a problem if the woofer in the coax works in bass frequencies also. In a traditional 2-way design the drivers are separated and will also normally sit in different enclosures, thus the same doppler shift shouldn't happen. But there are other problems with crossing a woofer very high and why I never do that myself.
 

fpitas

Master Contributor
Forum Donor
Joined
Jul 7, 2022
Messages
9,885
Likes
14,213
Location
Northern Virginia, USA

AdamG

Helping stretch the audiophile budget…
Moderator
Forum Donor
Joined
Jan 3, 2021
Messages
4,750
Likes
15,744
Location
Reality
There was recently a very similar thread about it:

Duplicate Threads merged here. Thank you for your assistance with keeping our house organized Sir! :cool:
 

Bjorn

Major Contributor
Audio Company
Forum Donor
Joined
Feb 22, 2017
Messages
1,313
Likes
2,602
Location
Norway
I'll come back with more measurements in a separate thread, but here's a first on-axis measurement of the KEF Q350 in near field.

1/48 oct smoothing:
KEF Q350 20 cm distance on axis.jpg


1/6 Oct smoothing:
KEF Q350 20 cm distance near field 1 to 6 oct.jpg



Seems to correspond quite well with Stereophile's measurement:
 

Ron Texas

Master Contributor
Forum Donor
Joined
Jun 10, 2018
Messages
6,249
Likes
9,389
The difference is in dispersion. Also, the LS50 is a coaxial design although I'm not sure where that shows up in the measurements.
 

HarmonicTHD

Major Contributor
Joined
Mar 18, 2022
Messages
3,326
Likes
4,837
I'll come back with more measurements in a separate thread, but here's a first on-axis measurement of the KEF Q350 in near field.

1/48 oct smoothing:
View attachment 326369

1/6 Oct smoothing:
View attachment 326370


Seems to correspond quite well with Stereophile's measurement:
Have a look at Amir’s measurements I posted above. Pretty close too at first glance. Still not convinced it is inherent to the coax-also see above.
 
Top Bottom