• WANTED: Happy members who like to discuss audio and other topics related to our interest. Desire to learn and share knowledge of science required as is 20 years of participation in forums (not all true). There are daily reviews of audio hardware and expert members to help answer your questions. Click here to have your audio equipment measured for free!

KEF LS50 Meta Review (Bookshelf Speaker) Video

abdo123

Master Contributor
Forum Donor
Joined
Nov 15, 2020
Messages
5,675
Likes
5,510
Location
Brussels, Belgium
Well -sue me but I can't hear any difference between the originals and the LS50 Metas. I like my originals in Racing Red, so I kept those. I don't believe a piece of rubber makes a big difference, and my listening session confirmed my suspicion. I think KEF is trying to drive the same upgrade cycle as smartphones go through, experimenting with marginal innovation and price and see how that drives buyer behavior.

The key to me is -as has been incessantly stated- to pair them with a good, musical sub, but also very importantly to set the sub crossover anywhere from 60-80Hz to let the LS50 focus on what they do best. Different amps will have different slopes, so experiment with whatever sound best for you. I'd say if you're a basshead you may want 2 good subs and set things at 120Hz or so, but then you'll have to set up things very carefully. The rewards will be huge though.

I'd venture to say if you carefully balance any LS50 with optimal sub crossover, placement etc for your room, you'll get total reference sound. Prolly much better and easier (and at a fraction of the cost) of reference full-range floorstanders (which are near impossible to set up optimally in anything other than a purpose-designed, large room).
How far do you sit from your speakers?
 

pablolie

Senior Member
Forum Donor
Joined
Jul 8, 2021
Messages
372
Likes
588
Location
bay area, ca
BoredErica
> There were multiple changes coming into Meta, not just the metamaterial. Are you saying Kef took 9 years to refresh the speaker
> and all they did was ...
The big news -as the name implies- was indeed the meta-material. As to the rest, for the most part we're all left to guess. :)

> Audio in general is like that. The difference between half decent speakers aren't HUGE. ...
I completely concur. The measurements of both versions are actually extremely close - to the point one has to objectively wonder if they are indeed discernable to the human ear. To me, in the same system -which was carefully set up using Dirac info- they were not, and I decided to keep the original ones for pure esthetics. :) I sold the Metas to a friend.

> What makes a sub musical vs good but not musical?
The two mean the same, I think. Control over the limited range is probably what you'd measure. We all know setup is key to get balanced performance with a sub, and we have all heard systems where that is not remotely so... :)

abdo123:
> How far do you sit from your speakers?

These days, for true listening, around 7ft and straight in the middle with the speakers pointed at me. The speakers are about 6ft apart. The sub is where it was calibrated to go. Cutoff is 70Hz, but the LS50 start to slightly come to life at 50Hz when I try test tones, we all know the rolloff is not abrupt. It's clear the LS50 work much harder with that 50Hz test tone when I turn off the sub crossover. Even though I never listen too loud (I only do enjoyable, dedicated music listening with jazz and classical, with other music I don't think too much about it), I imagine that really reduces distortion with the LS50. To me - that did a huge difference in my enjoyment of the LS50, offloading the lower bass. I wonder if that's why I didn't hear a difference with the Metas, because they sound better in a full range use case? BTW my friend has a bigger room than mine, doesn't use a sub and is super happy with the Metas. There's no doubt they are fantastic speakers.
 

groovybassist

Member
Forum Donor
Joined
Dec 11, 2020
Messages
89
Likes
116
Location
Wendell, NC
BoredErica
> There were multiple changes coming into Meta, not just the metamaterial. Are you saying Kef took 9 years to refresh the speaker
> and all they did was ...
The big news -as the name implies- was indeed the meta-material. As to the rest, for the most part we're all left to guess. :)

> Audio in general is like that. The difference between half decent speakers aren't HUGE. ...
I completely concur. The measurements of both versions are actually extremely close - to the point one has to objectively wonder if they are indeed discernable to the human ear. To me, in the same system -which was carefully set up using Dirac info- they were not, and I decided to keep the original ones for pure esthetics. :) I sold the Metas to a friend.

> What makes a sub musical vs good but not musical?
The two mean the same, I think. Control over the limited range is probably what you'd measure. We all know setup is key to get balanced performance with a sub, and we have all heard systems where that is not remotely so... :)

abdo123:
> How far do you sit from your speakers?

These days, for true listening, around 7ft and straight in the middle with the speakers pointed at me. The speakers are about 6ft apart. The sub is where it was calibrated to go. Cutoff is 70Hz, but the LS50 start to slightly come to life at 50Hz when I try test tones, we all know the rolloff is not abrupt. It's clear the LS50 work much harder with that 50Hz test tone when I turn off the sub crossover. Even though I never listen too loud (I only do enjoyable, dedicated music listening with jazz and classical, with other music I don't think too much about it), I imagine that really reduces distortion with the LS50. To me - that did a huge difference in my enjoyment of the LS50, offloading the lower bass. I wonder if that's why I didn't hear a difference with the Metas, because they sound better in a full range use case? BTW my friend has a bigger room than mine, doesn't use a sub and is super happy with the Metas. There's no doubt they are fantastic speakers.
Lots of info on the LS50 Meta changes in KEF’s white paper.

https://us.kef.com/pub/media/wysiwyg/ls50/20-KEF-LS50-Collection-WhitePapers.pdf
 

pablolie

Senior Member
Forum Donor
Joined
Jul 8, 2021
Messages
372
Likes
588
Location
bay area, ca
Thanks - we all know that driving upgrade cycles is a huge motivator in technology (I work in Silicon Valley in High Tech). I also spend a lot of time writing white papers, but it's not in consumer electronics at all, so I am not claiming more familiarity with the topic we're discussing than any other of the very smart contributors that bless this forum.
I found the white paper a great read. I do however question stuff like Figure 9, where they show a comparison between what seem identical parts, just one is fully assembled and the other is not. The difference seems to be in the metal used at the base, and whether it's for improvement or for cost savings is anyone's guess. :)
We all know there are 2 things that drive successful products' development: (a) Milk more profit from the same functional product via cost reduction, and (b) genuine innovation to stay ahead of the inevitable competition a successful product will attract.To even remotely think that KEF was only and exclusively motivated by (b) is a naive assumption. :)
 

Hapo

Active Member
Joined
Jan 11, 2022
Messages
225
Likes
96
...now I have a grasp on wtf all those lines mean...
 

KMO

Senior Member
Joined
Aug 9, 2021
Messages
442
Likes
590
I do however question stuff like Figure 9, where they show a comparison between what seem identical parts, just one is fully assembled and the other is not. The difference seems to be in the metal used at the base, and whether it's for improvement or for cost savings is anyone's guess. :)
The difference being pointed out is increased duct size, needed to make the metamaterial work better, and consequent motor redesign. It's talked about from 27:00 in the video.

Basically, the old tweeter had a 12mm duct, and the new one has a 19mm duct, and the magnet system has been narrowed and made more linear while they were at it.
 

pablolie

Senior Member
Forum Donor
Joined
Jul 8, 2021
Messages
372
Likes
588
Location
bay area, ca
The difference being pointed out is increased duct size, needed to make the metamaterial work better, and consequent motor redesign. It's talked about from 27:00 in the video.
Basically, the old tweeter had a 12mm duct, and the new one has a 19mm duct, and the magnet system has been narrowed and made more linear while they were at it.
Well, once again the only thing I can say is that I got a pair, but stayed with the old ones. I didn't hear any difference at all, and visually preferred the old color. It's also peculiar that KEF hasn't found it necessary to meddle around with the coaxial stuff in their higher end products (although of course they always listed the LS50 along those as reference) ever, or add "metamaterial" to them. :) To me it sounds like a clever ploy to drive an upgrade cycle with a very successful products. And more power to them. :)
 

KMO

Senior Member
Joined
Aug 9, 2021
Messages
442
Likes
590
It's also peculiar that KEF hasn't found it necessary to meddle around with the coaxial stuff in their higher end products (although of course they always listed the LS50 along those as reference) ever, or add "metamaterial" to them. :)
Every higher-end product released or revised since the LS50 Meta came out has had the metamaterial absorption technology. ;)
 

pablolie

Senior Member
Forum Donor
Joined
Jul 8, 2021
Messages
372
Likes
588
Location
bay area, ca
Every higher-end product released or revised since the LS50 Meta came out has had the metamaterial absorption technology. ;)
Which is synonymous with *none*. :) You'd think they'd introduce groundbreaking innovative technologies at the top end first... :-D
 

pablolie

Senior Member
Forum Donor
Joined
Jul 8, 2021
Messages
372
Likes
588
Location
bay area, ca
And if you don't believe my comparison in my own room, check the measurements of either in this very web site, side by side. Oddly one is recommended "very close to ideal", the other -with identical behavior- not, which is bizarre, but hey - what I love about this website is it allows us to make up our own mind and reject the subjective and see the measurements. Both are stellar speakers - and their in-room response {as well as pretty much every measurement) is damn-near-effing identical. If anything the original looks a tad more linear there. And yes, there are other aspects to consider. But let's be honest - anyone that claims to hear a clear difference...
 

Attachments

  • ls50compa.png
    ls50compa.png
    936.4 KB · Views: 51
Last edited:

BoredErica

Addicted to Fun and Learning
Joined
Jan 15, 2019
Messages
529
Likes
702
Location
USA
You just need to bother spending the time going to Spinorama github, open up a tab for meta and one for ls50 and tab back and forth.
On%20Axis.jpg

On%20Axis.jpg
Meta is flatter here in that the dips and peaks are less but also the overall slope of the curve is closer to 0 vs LS50 which is going upwards.
CEA2034.jpg

CEA2034.jpg
LW is closer to on axis. Early reflections DI and even more so sound power DI is smoother with Meta. It is going up and down on Ls50.
SPL%20Horizontal%20Normalized.jpg

SPL%20Horizontal%20Normalized.jpg
Look at how much LS50 goes up past 4khz before coming back down. 1khz is more crushed too. Better directivity allows better EQing. This alone ends the debate.
Estimated%20In-Room%20Response.jpg

Estimated%20In-Room%20Response.jpg
Dispersion contours normalized:
SPL%20Horizontal%20Contour%20Normalized.jpg

SPL%20Horizontal%20Contour%20Normalized.jpg
These measurements are too different for you to start saying people who hear a difference are just suffering from placebo. Trained listeners can hear distortions many, many, many db below the average person. There are even tracks designed to tease out differences in tonality. If you think that peak of LS50 at 4-5khz is too high Q to be audible, then look at the dip on the LS50 from 400-600hz with a difference of 2db between the two.

You might not hear a difference and that doesn't mean your ears are broken. That's a perfectly reasonable statement to make. But you didn't want to stop there. All this time writing white papers and you don't know you're not supposed to compare two curves with different axises.

ok I'm done arguing with people who can't be bothered to read past the first graph in each review.
 
Last edited:

pablolie

Senior Member
Forum Donor
Joined
Jul 8, 2021
Messages
372
Likes
588
Location
bay area, ca
You just need to bother spending the time going to Spinorama github, open up a tab for meta and one for ls50 and tab back and forth.
On%20Axis.jpg

On%20Axis.jpg
Meta is flatter here in that the dips and peaks are less but also the overall slope of the curve is closer to 0 vs LS50 which is going upwards.
CEA2034.jpg

CEA2034.jpg
LW is closer to on axis. Early reflections DI and even more so sound power DI is smoother with Meta. It is going up and down on Ls50.
SPL%20Horizontal%20Normalized.jpg

SPL%20Horizontal%20Normalized.jpg
Look at how much LS50 goes up past 4khz before coming back down. 1khz is more crushed too. Better directivity allows better EQing. The above to sets of curves are most definitely NOT THE SAME at all. This entire debate is over, this is the killshot.
Estimated%20In-Room%20Response.jpg

Estimated%20In-Room%20Response.jpg
Dispersion contours normalized:
SPL%20Horizontal%20Contour%20Normalized.jpg

SPL%20Horizontal%20Contour%20Normalized.jpg
These measurements are too different for you to start saying people who hear a difference are just suffering from placebo. Trained listeners can hear distortions many, many, many db below the average person. If you think that peak of LS50 at 4-5khz is too high Q to be audible, then look at the dip on the LS50 from 400-600hz with a difference of 2db between the two.

You might not hear a difference and that doesn't mean your ears are broken, but if you're willing to bet money nobody else can hear a difference you're going way too far. All this time writing white papers and you don't know you're not supposed to compare two curves with different axises.

ok I'm done arguing with randos on the internet who are too lazy to open up a few images in new tabs on spinorama github, or scroll down a review more than the first graph in each review.
They are extremely visibly not. Thanks for the effort in further making the point. :)

PS: I never argue. You may have the last word despite the fact everything you posted points at minimal deltas that can be explained by slight differences in measurement setup or conditions. Come on. They are THE SAME.
 
Last edited:

KMO

Senior Member
Joined
Aug 9, 2021
Messages
442
Likes
590
Which is synonymous with *none*. :) You'd think they'd introduce groundbreaking innovative technologies at the top end first... :-D
Given product lifecycles, that would mean doing nothing visible for 10 years then suddenly putting 10 years of work in one product. And all those simultaneously-revealed developments would then be competing for attention/marketing clout. Seems fairly obvious why that's not the approach.

But sure, they do have to keep in mind that it's bad if "top end" products do end up seeming too dated. They've occasionally done refreshes like the 207/2 or the revised Muon.
 

thewas

Major Contributor
Forum Donor
Joined
Jan 15, 2020
Messages
4,171
Likes
9,148
You may have the last word despite the fact everything you posted points at minimal deltas that can be explained by slight differences in measurement setup or conditions. Come on. They are THE SAME.
Nope, all sources that have measured both confirm those on-axis, directivity and distortion differences, here for example Soundstage
and my own measurements (I own both)
I own the original ones now almost 10 years and without EQ they are unlistenable for me, but with EQ the differences shrink and they become also nice, although the Meta are still better.

PS: Interesting also to read the white paper which shows all the improvements in comparison to the old ones, which are much more than just the metamaterial rear wave tweeter absorber https://assets.kef.com/pdf_doc/LS50Meta/KEF-LS50Meta-LS50WirelessII-WhitePaper.pdf
 

pablolie

Senior Member
Forum Donor
Joined
Jul 8, 2021
Messages
372
Likes
588
Location
bay area, ca
I own the original ones now almost 10 years and without EQ they are unlistenable for me
I then question why you made the mistake to buy unlistenable speakers in the first place. And live with unlistenable speakers 10 years.
 

KMO

Senior Member
Joined
Aug 9, 2021
Messages
442
Likes
590
Interesting also to read the white paper which shows all the improvements in comparison to the old ones, which are much more than just the metamaterial rear wave tweeter absorber
I actually think the MAT is probably one of the least significant changes in terms of actual result. It strikes me more as an engineering research project, and something they're going to be trying out more going forward. But it makes for good marketing - highlighting their technical chops. And it's ultimately cheaper to produce, I bet - just a few simply molded plastic parts. Win-win-win.

But I reckon the still-highly-technical but just "incremental"-looking stuff done to the existing systems probably matters more. The tweeter gap damper adapted from the R series, for example - they can highlight that change on the total output. No equivalent graph for the MAT...

And the ferrite magnet system improvements are edging them closer to the low-distortion neodymium Reference motors, but they'll still be a long way off. Don't make the mistake of thinking the LS50 Meta is somehow better than the Reference because it's newer. The ridiculously-low distortion Amir measured from the Reference 4c comes at a cost point the LS50 Meta isn't competing at.
 

thewas

Major Contributor
Forum Donor
Joined
Jan 15, 2020
Messages
4,171
Likes
9,148
I then question why you made the mistake to buy unlistenable speakers in the first place. And live with unlistenable speakers 10 years.
Better is the enemy of good, we get used to correct tonality only when we listen to it and it is a process that doesn't happen in a day and I use EQ since a long time with them and except the tonal tuning the other qualities are great, still use them at my desktop system today.
 

thewas

Major Contributor
Forum Donor
Joined
Jan 15, 2020
Messages
4,171
Likes
9,148
I actually think the MAT is probably one of the least significant changes in terms of actual result. It strikes me more as an engineering research project, and something they're going to be trying out more going forward. But it makes for good marketing - highlighting their technical chops. And it's ultimately cheaper to produce, I bet - just a few simply molded plastic parts. Win-win-win.

But I reckon the still-highly-technical but just "incremental"-looking stuff done to the existing systems probably matters more. The tweeter gap damper adapted from the R series, for example - they can highlight that change on the total output. No equivalent graph for the MAT...

And the ferrite magnet system improvements are edging them closer to the low-distortion neodymium Reference motors, but they'll still be a long way off. Don't make the mistake of thinking the LS50 Meta is somehow better than the Reference because it's newer. The ridiculously-low distortion Amir measured from the Reference 4c comes at a cost point the LS50 Meta isn't competing at.
Well analysed, like you say the Meta makes probably the smallest change to its acoustic output. Now about the 4c, it cannot be compared of course in terms of distortion due to being a 3-way and having also several woofers although above 200 Hz the Meta does quite well, especially considering its size.
 
Top Bottom