• 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!

KEF R3 meta Measurements

Nuyes

Active Member
Forum Donor
Reviewer
Joined
Jun 8, 2022
Messages
218
Likes
3,598
Location
South Korea
Hi,
I measured the KEF R3 meta loudspeakers this time.

These speakers have just been released, but I was able to get a pre-order sample (a pair) from an audiophile in Korea.

00.JPG











Impedance

01.png
02.png









Frequency response



03.png


I measure HF responses using a 5ms window gating.
Therefore, the crossover frequency of this speaker was located lower than I could measure, so I could not use the 'Merging Nearfield + Farfield' technique.


It has a similar LF roll-off to its predecessor.





04.png


Here's a comparison of 'ported vs sealed' responses.
(This was done with nearfiled measurements, with baffle step compensation applied).





Directivity



05.png
06.png
07.png
08.png





Wow. That's really cool.

I'm noticing a huge improvement in the 15~20k compared to before.

On the outside, it looks like the exact same product as its predecessor, but the directivity definitely seems to be a step up.

While this may not be important to all users, I can't help but admire KEF's passion and skill in achieving performance improvements despite the seemingly identical design.

(Perhaps it's a fine-tuning of the lemon squeezer on tweeter?)




Beamwidth


09.png
10.png



It looks as complete as anything you'll see in this price range.
(In South Korea, it's currently available for around $1,700 to $1,800).










Polar plot




11.png
12.png




This loudspeaker shows almost perfectly smooth attenuation, not only in the front, but also in the sides and back.




Total Harmonic Distortion


I am representing THD relative data as CHD.
This is the percentage of each HD component based on the average SPL from 200 Hz to 10 kHz.

13.png
14.png
15.png
16.png



You can hear a very clean response at 85dB SPL and even at 95dB SPL.
It's flawless.




Multitone test

17.png
18.png


Multitone distortion of -40 to -50 dB were measured.









I tried removing the measurement signal below 80 Hz and measuring it.
This was done to see how much the effects of MD from the woofer's high displacement motion and ports could be reduced when used with a subwoofer.



19.png











And if anyone remembers, this product also had the "shadow flare" issue that was featured in Erin's Audio Corner.
So I measured how much the MD was changed by this as well.
(The effect on FR can be seen below).

20.png








MD for each SPL.



21.png
22.png




Despite the increase in output, there is very little increase in Multitone Distortion.
This is a strong defense against distortion.




Compression test


23.png



This test was performed with Multitone.
It shows a compression of only 0.5 dB even at 96 dB SPL.
Very powerful performance.






Here's a look at compression testing from the other side.

These were performed at 0.1 m and 1 m, respectively, and measured with a sine sweep.



24.png
25.png



We can see that the response of the woofer increased with the increase in SPL at 45Hz.
This means that the internal air pressure along the port did not remain linear.

In other words, above a certain sound pressure, the ports in this speaker lose a little bit of their effectiveness.

(All SPL values are based on output at 1kHz @1m)






Grill test


26.png



The grille is affected by a very high Q value.







Shadow flare


27.png



I measured the impact of Shadow flare, which we introduced in EAC.
(I don't own this loudspeaker, so I wasn't able to push the part very hard).

You can see that there is a dip centered around 1 kHz.
(Pressing in the Shadow flare will correct this).




28.png






Finally, the FR deviation comparison between the two samples.

29.png






A very personal thought.

KEF's precise directional control is truly amazing.

Its predecessor was good enough, but it's even better than I expected.


If I had to pick a downside...
The bass is weak.
Also, due to the compression at high SPL, I think it would be a bit too much for me to enjoy a movie without a subwoofer.

Of course, I was just nitpicking because it was so great. :)
(I envy the owner.)
 
OP
Nuyes

Nuyes

Active Member
Forum Donor
Reviewer
Joined
Jun 8, 2022
Messages
218
Likes
3,598
Location
South Korea
Always thanks for your work! New R3 is surely upgraded ... quite impressive (especially ditorstion graphs). I can notice many good points compare to before R3's measurements you made. I want to get a new R3!

Fantastic upgrade!
Once again, KEF has exceeded our expectations.
(Honestly, I doubted this was possible without going beyond the Reference series.

The fine tuning for the class of each series is amazing.
 

delta76

Major Contributor
Joined
Nov 27, 2021
Messages
1,653
Likes
2,599
Thank you for your work. I don't understand those graphs, but in layman terms, they are even better than Kef R3? how much better?
 
OP
Nuyes

Nuyes

Active Member
Forum Donor
Reviewer
Joined
Jun 8, 2022
Messages
218
Likes
3,598
Location
South Korea
Thank you very much for those measurements, could you maybe plot some direct comparison plots to the old R3, for example for 96 dB MD relative to fundamental and 6/12 dB horizontal bandwidth?
I am leaving on a business trip for a few days.
I'll check it when I get back.

Thank you. :)
 
OP
Nuyes

Nuyes

Active Member
Forum Donor
Reviewer
Joined
Jun 8, 2022
Messages
218
Likes
3,598
Location
South Korea
Thank you for your work. I don't understand those graphs, but in layman terms, they are even better than Kef R3? how much better?
KEF has made a successful upgrade.
It's cleaner, has a flatter directivity, and delivers the sound source more accurately.
 

Mario Sanchez

Active Member
Joined
Jan 11, 2021
Messages
161
Likes
277
It's a bit unfortunate that the beamwidth "step" is still there from the last gen (~+-60 for midrange before XO, ~+-45 for tweeter after XO), but the distortion looks pretty clean for size. Regarding the shadow flare, did you find it difficult to push the part fully in?
Oh and as an afterthought, you wouldn't also happen to have a R2 Meta lying around...right?
 

Vacceo

Major Contributor
Joined
Mar 9, 2022
Messages
2,757
Likes
2,914
Thank you, Nuyes, first of all. I'd like to ask you about your comment on bass extension: how well does it work on that area considering the limitations of a bookshelf?

On the con of the SPL, would you say a tower of the same family (R7 for example) could do it with better ease being the R3 more suited for sides or rears?
 
OP
Nuyes

Nuyes

Active Member
Forum Donor
Reviewer
Joined
Jun 8, 2022
Messages
218
Likes
3,598
Location
South Korea
It's a bit unfortunate that the beamwidth "step" is still there from the last gen (~+-60 for midrange before XO, ~+-45 for tweeter after XO), but the distortion looks pretty clean for size. Regarding the shadow flare, did you find it difficult to push the part fully in?
Oh and as an afterthought, you wouldn't also happen to have a R2 Meta lying around...right?
This is not my property, so I can't push it hard.
(to avoid any possible internal damage).

And in a few weeks my studio will be receiving the R6 meta.
 

pjn

Active Member
Joined
Aug 28, 2021
Messages
178
Likes
172
This is not my property, so I can't push it hard.
(to avoid any possible internal damage).

And in a few weeks my studio will be receiving the R6 meta.
The R6 is an interesting option - the R3 (pre-meta) didn't sound much different from an LS50 meta as a center, although with slightly better bass
 

Daka

Senior Member
Joined
Oct 6, 2021
Messages
334
Likes
220
This is not my property, so I can't push it hard.
(to avoid any possible internal damage).

And in a few weeks my studio will be receiving the R6 meta.
R6 is very good, and surprised me how much better compared to R2C. Recommended and can’t wait for your measurements of it.
 

regan

Active Member
Joined
Jun 14, 2022
Messages
197
Likes
86
What preference score could be deduced from these measurements?
 
OP
Nuyes

Nuyes

Active Member
Forum Donor
Reviewer
Joined
Jun 8, 2022
Messages
218
Likes
3,598
Location
South Korea
For a lot of people, this with a sub could truly be endgame. And they're probably gonna go on sale from time to time, just like their predecessor. Wow.
I completely agree.
There are many reasons why people like their respective loudspeakers, but when it comes to price/performance, it's hard to beat KEF.
At least if you're aiming for ideal performance.
 
OP
Nuyes

Nuyes

Active Member
Forum Donor
Reviewer
Joined
Jun 8, 2022
Messages
218
Likes
3,598
Location
South Korea
Thank you, Nuyes, first of all. I'd like to ask you about your comment on bass extension: how well does it work on that area considering the limitations of a bookshelf?

On the con of the SPL, would you say a tower of the same family (R7 for example) could do it with better ease being the R3 more suited for sides or rears?
If we accept the limitations of this speaker as a bookshelf, the bass looks good.

However, I think we need to be a little more conservative than the product specs indicate.
 

tehas

Member
Joined
Oct 22, 2022
Messages
42
Likes
26
Forgive me if this is off topic - I'm curious about the Klippel measurement system. It doesn't look like the same setup as Amir/Erin use - this is not the NFS, correct? Can the software be used with off the shelf/non-Klippel measurement hardware as well? What are the license requirements?
 
OP
Nuyes

Nuyes

Active Member
Forum Donor
Reviewer
Joined
Jun 8, 2022
Messages
218
Likes
3,598
Location
South Korea
Forgive me if this is off topic - I'm curious about the Klippel measurement system. It doesn't look like the same setup as Amir/Erin use - this is not the NFS, correct? Can the software be used with off the shelf/non-Klippel measurement hardware as well? What are the license requirements?
Yes, that's right.
I use Klippel Distortion Analyzer, not NFS.

The Klippel software you are referring to is called dB-Lab, and it is free for the viewer only.

However, if you want to use it to make measurements or do special operations like directivity plots, you need hardware and a license.
 
OP
Nuyes

Nuyes

Active Member
Forum Donor
Reviewer
Joined
Jun 8, 2022
Messages
218
Likes
3,598
Location
South Korea
What preference score could be deduced from these measurements?
I don't have NFS, so I can't get the full Spinorama data.
Therefore, I'm focusing more on the other characteristics (ex: MD, Compression) that are worth noting.

It's a very difficult thing to score.
Especially when it comes to distortion, because the actual auditory feel of each distortion is different depending on its source and degree.

The same % of distortion can be annoying in different ways depending on what causes it.

There are even issues that don't show up at all in THD measurements.

For example...
Turbulent noise from ports, air leaks in the speaker box, and vibration from internal components are random noise and will not be detected by a THD measurement.

Also, intermodulation distortion in the speaker unit, distortion due to the Doppler effect, etc. are never detected by THD measurements.

However, these are important factors when we judge speaker performance by our 'ears'.
 
Top Bottom