KEF's new Metamaterial Absorption Technology

[tvrgeek]

My point is a lot of comments, starting with clarifying how KEF advertising is specious at best, but they are a respected company and have made great products. All this thread has done is try to justify their claims but seems no one here commenting has heard them.

104s were the first actual HI-FI speakers I ever heard. I grew up wit AR2's and JBLs not knowing any better. But my last Kef's were Q1's and I was very disappointed in them. I wish I could, but not traveling across states during COVID just out of curiosity.

 

[MZKM]

KEF LS50 Meta NRC measurements here:

https://www.soundstagenetwork.com/i...&catid=77:loudspeaker-measurements&Itemid=153

Disappointing, it only looks like a crossover redesign. Besides a bit less bass (absorbing cabinet vibration), I don’t see any indication that the Meta material is doing anything.

 

[thewas]

Disappointing, it only looks like a crossover redesign. Besides a bit less bass (absorbing cabinet vibration), I don’t see any indication that the Meta material is doing anything.

Metamaterial is used for the absorption of the tweeter so has nothing to do with the bass and from the shown plots it thus cannot be seen if there are distortion differences as in the tweeter region the plot is on both under the graph limits.

 

[Zvu]

@MZKM You will probably miss the differences if you look at those few measurements. There were problems with old LS50 beyond crossover.

Tweeter had obvious (cavity ?) resonance around 2kHz. Given the shallow slopes of old ls50, that resonance wasn't well supressed in level. Even with completely redesigned crossover with flat frequency response some people characterised kef tweeter as harsh.

Meta should take care of tweeter resonances and you'd be surprised how the same drivers and cabinet can sound with properly designed crossover.

They've kept the same price, giving better objective performance (easily seen even in rudimentary measurements such are LW and horizontal off axis frequency response) so i don't think they deserve bashing.

Edit: @thewas_ beat me to it

 

[thewas]

(very good) NRC measurements of the new passive LS50 Meta
https://www.soundstagenetwork.com/i...ef-ls50-meta-loudspeakers&catid=77&Itemid=153
KEF seems to have successfully corrected the FR problems of the first passive one
https://www.soundstagenetwork.com/i...&catid=77:loudspeaker-measurements&Itemid=153
rest looks as excepted very similar, "unfortunately" already the distortion measurements of the old tweeter were beyond the plot scale so we cant see the differences of the new one with the META absorption material, also as per NRC standard unfortunately no decay spectrum measurements.

The corresponding review and interesting listening comparisons https://www.soundstagehifi.com/index.php/equipment-reviews/1502-kef-ls50-meta-loudspeakers
Sure, reviews always should be taken with a grain of salt but what Doug Schneider writes corresponds quite well with the measurements and also personal experiences.

 

[Ilkless]

The corresponding review and interesting listening comparisons https://www.soundstagehifi.com/index.php/equipment-reviews/1502-kef-ls50-meta-loudspeakers
Sure, reviews always should be taken with a grain of salt but what Doug Schneider writes corresponds quite well with the measurements and also personal experiences.

I'm sure they aren't really audible, but there's a high-Q 1.3kHz peak on the far off-axis and several high-Q disturbances around 2kHz adjacent to each other. I wonder what is the source of that. The absorber?

 

[thewas]

I'm sure they aren't really audible, but there's a high-Q 1.3kHz peak on the far off-axis and several high-Q disturbances around 2kHz adjacent to each other. I wonder what is the source of that. The absorber?

Well observed, the old LS50 doesn't seem to have them, would be also curious to find out about them.

 

[renevoorburg]

I've combined the LS50 vs the LS50 Meta measurements from https://www.soundstagenetwork.com/ in one image:

 

[Deleted member 14468]

As the person who wrote the review, I can say that the speakers are certainly similar, but definitely not the same. The two main differences come in the upper midrange, which I think correlates well with the measurements -- there are significant-enough differences in the 1kHz-5kHz range to be easily audible -- as well as in the way the speaker gets the sound "out of the box," which isn't so easy to read in the graphs. But there are smaller differences in the bass and lower midrange, too. I don't really need to say too much more -- my review is long enough already:

https://www.soundstagehifi.com/index.php/equipment-reviews/1502-kef-ls50-meta-loudspeakers

Doug Schneider
www.SoundStage.com

 

[Ilkless]

The Soundstage listening window weightage is different from the Olive one it seems, but they are comparable between each other. Red lines denote +/- 1.5dB bounds. Yellow areas represent areas with distinctly more ripple than the rest of the band. IMHO they sum up the improvement KEF has made here.

Meta:

Original:

Both are within +/- 1.5dB bounds within the listening window. The Meta is within +/- 1dB except for a couple of minuscule spots probably within experimental error. Note that the original LS50 has a 2dB bump from 500Hz to 1kHz, and a 3dB bump from 1.5kHz to 7kHz, relative to a reference SPL of 83dB. Ironically, the original may have been a victim of its own success. It is remarkably flat except for the 2 broad low-Q bumps above 500Hz. I suspect that more ripple across the band would have made the 2 bumps less distinct to both the eye and ear. In fact the discontinued R100 that was released at the same time as the LS50 was precisely that IME. More ripple but no areas of broad, distinct low-Q deviation, and I had found it preferable to the original LS50 in the exact same spot, in the same room, when I heard them before measurements were released.

Furthermore, the Meta is all dips from the baseline. The original is all peak. Dips being less audible than peaks of course.

edit: R100 (discontinued) for comparison.

 

[Trouble Maker]

I found what's at the middle of the maze... it's MJ.

 

[hardisj]

I've combined the LS50 vs the LS50 Meta measurements from https://www.soundstagenetwork.com/ in one image:
View attachment 94487

This actually looks quite good and, at least to me, would justify the purchase of the new ones if I were on the fence.

Though, I am surprised to see a LF difference. Hard to get a hold on exactly, but the Meta looks to lose about 3dB below 90Hz compared to the original. Is this measurement-borne (and not really accurate of the true response)? If this is "real", then that means more changed than what I thought.

 

[Ilkless]

This actually looks quite good and, at least to me, would justify the purchase of the new ones if I were on the fence.

Though, I am surprised to see a LF difference. Hard to get a hold on exactly, but the Meta looks to lose about 3dB below 90Hz compared to the original. Is this measurement-borne (and not really accurate of the true response)? If this is "real", then that means more changed than what I thought.

Keep in mind response below 100Hz in the NRC should be disregarded. Their anechoic chamber is not effective below that and they use some sort of compensation function that ends up off for many speakers.

 

[hardisj]

Keep in mind response below 100Hz in the NRC should be disregarded. Their anechoic chamber is not effective below that and they use some sort of compensation function that ends up off for many speakers.

That is what I was thinking. But, I would hope their own results would be directly comparable. And that's the part that catches my attention. Either the measurements weren't under the same conditions or the speaker really is different.

 

[Ilkless]

That is what I was thinking. But, I would hope their own results would be directly comparable. And that's the part that catches my attention. Either the measurements weren't under the same conditions or the speaker really is different.

I think it's a combination of the speaker coupling with the room <100Hz differently (due to different bass alignment), and how the (unreliable) compensation function then interacts with that. A lot of confounding variables, that can swing in both in directions. I forgot my source, but this was my understanding of it.

It's anechoic down to 80Hz but let's say 100Hz to be safe. They found another chamber anechoic down to 20Hz, then measured the difference between both chambers as the compensation curve. It is then applied to the measurement of the DUT (measured at 2m with a full-range sweep despite the non-anechoic environment <100Hz) at the NRC. But different speakers couple differently to the remaining modes and nulls <100Hz. So applying a one-size-fits-all compensation curve to it is bound to be a confounding variable in both directions (ie. SPL can be both higher or lower than it actually is in either case and we can't isolate the variations).

I personally wish they just measured bass nearfield and spliced it, rather than a full-range sweep + compensation curve at 2m.

 

[Deleted member 14468]

It's funny how stories get spun -- and spun out of control if someone doesn't interject.

The LS50 and LS50 Meta were measured 9 years apart, but the conditions are the roughly the same -- the electronics and room and platform haven't changed. Yes, it is true that the chamber is unreliable below 100Hz, but I think something about the bass alignment has changed. The old LS50 is trailing off above 100Hz, the new one more sharply after. These are questions for the designers, really.

As far as the correction. Back when, a combination of freefield measurements and measurements from another chamber were used to apply a "general" correction curve. Anyone who has worked with these types of corrections knows that it's never exact and that it varies from speaker to speaker. So below 100Hz it's ballpark -- but, still, there appears to be something different between the old and new LS50, and in listening, I sensed a subtle difference.

Doug Schneider
SoundStage!

 

[Ilkless]

It's funny how stories get spun -- and spun out of control if someone doesn't interject.

The LS50 and LS50 Meta were measured 9 years apart, but the conditions are the roughly the same -- the electronics and room and platform haven't changed. Yes, it is true that the chamber is unreliable below 100Hz, but I think something about the bass alignment has changed. The old LS50 is trailing off above 100Hz, the new one more sharply after. These are questions for the designers, really.

As far as the correction. Back when, a combination of freefield measurements and measurements from another chamber were used to apply a "general" correction curve. Anyone who has worked with these types of corrections knows that it's never exact and that it varies from speaker to speaker. So below 100Hz it's ballpark -- but, still, there appears to be something different between the old and new LS50, and in listening, I sensed a subtle difference.

Doug Schneider
SoundStage!

Yeah, but my point is that a different bass alignment would couple with the non-anechoic <100Hz of the room differently, and that the fixed correction curve might not correct for both to the same degree, or in the same direction (even though the curve as I understand is most optimised for monopoles which both are). It could very well be different, but I would not draw any conclusions as to the direction of the difference given the conditions. As you said it's a ballpark, and the ballpark has enough variance in these circumstances that there's a lot of overlap, so it could go both ways.

 

[audiopile]

I owned and enjoyed a set of LS-50's for a number of years (size was a hit with SWMBO'd -my homemade lead shot /spiked speaker stands -less so to her eyes). I liked them but they were a bit hard sounding to my ol' nun handles -eventually I sold them off and quit trying to find amps that fit them. This is still one of the rare speaker companies that actually maintains a engineering research lab. About a year before his passing I remember asking John Bowers what were the other speaker companies he paid attention to -his answer was instant "KEF of course" . BTW- lots of other folks in our hobby/obsession will tell you their product(s) is "IT" -all other competing products are shiehgt. Says something about the man . I expected a straight answer from him and got it.
Back to topic : the powered version with HDMI input is really tempting -once COVID is past I'm going to have to hit my local KEF dealer - these might be a lot closer to what I want in terms of frequency response.

 

[Ilkless]

I owned and enjoyed a set of LS-50's for a number of years (size was a hit with SWMBO'd -my homemade lead shot /spiked speaker stands -less so to her eyes). I liked them but they were a bit hard sounding to my ol' nun handles -eventually I sold them off and quit trying to find amps that fit them. This is still one of the rare speaker companies that actually maintains a engineering research lab. About a year before his passing I remember asking John Bowers what were the other speaker companies he paid attention to -his answer was instant "KEF of course" . BTW-this said about the man - lots of other folks in our hobby/obsession will tell you their product(s) is "IT" -all other competing products are shiehgt. Says something about the man . I expected a straight answer from him and got it.
Back to topic : the powered version with HDMI input is really tempting -once COVID is past I'm going to have to hit my local KEF dealer - these might be a lot closer to what I want in terms of frequency response.

I tried VERY hard to love the LS50 when it first came out. It was so far ahead in engineering to the usual "mainstream" hifi passives from Monitor Audio, Tannoy and the like back then, and they weren't afraid of substantiating it in their white papers, but the dispersion and frequency response just wasn't there yet.

The midrange was a tad forward, which I was still fine with. But the treble peak was ultimately overemphasising the leading edge on most of the brass on most recordings I tried it with and masking the trailing edge to a glaring (pun not intended) degree. There was significant sibilance as well. That I could not take.

 

[Deleted member 14468]

There was significant sibilance as well. That I could not take.

Interesting you had trouble with the sibilance. In comparisons with the new LS50 Meta, I noticed that too.

Doug
SoundStage!