KEF LS50 Wireless

[napilopez]

I’m curious about your measurements.

Unfortunately, whilst there are heaps of well done measurements for the passive LS50, there is a lack for the LS50w.

Stereo.de in Germany did some simplistic ones that also show a dip in the upper midrange, albeit somewhat lower than yours.

But what I find curious is that all the recent high end KEFs (new R series, Reference series) show exactly the same dip.

Given the LS50w has digital crossovers and EQ they could make the response as flat as they like.

But they have obviously chosen not to.

I’m assuming their version of a Spinorama says people prefer said dip, just as B&W believe people prefer a dip higher up and a tipped treble.

But I wish they wouldn’t.

I EQ the dip in my R3s out, but would prefer not to.

One other possible explanation is that they simply didn't think it was an audible enough issue to be worth correcting. In general dips are less audible than peaks, although one would think this is wide enough to be an issue. Still, some companies take a less is more approach and would rather make minimal tweaks to the frequency response. The fades away off axis so it seems like a directivity issue, but it seems to have little effect on the power/in-room response. Would probably be more audible in the nearfield though.

Not sure why I bother, but if you look at measurements for the LS50, there is no dip at 1K. There is a crossover peak at 2.5K.

Not sure where you saw the original LS50 doesn't have that dip? There's a clearly visible dip between 1Khz and 2k in Stereophile's measurements, as well as hifi-news' comparison of the LS50W with the passive model. Edit: Also Soundstage network

Note though how it seems KEF corrected the peaks on the LS50 with the LS50W.

 

[thefsb]

One other possible explanation is that they simply didn't think it was an audible enough issue to be worth correcting.

LS50W is a DSP-based system so wouldn't flattening it be very small R&D cost and zero per-unit cost?

 

[maty]

KEF Q100: It would take a very complex crossover to avoid it, I am afraid. Like the "V" SPL response, the important thing, to me, is like it is in the listening point. In my case, usually near field and speakers about 40º - 50º the SPL response is more interesting.

I have tried with different crossovers with LpsCAD. I made logical and informed decisions. I can not give up the 150 kHz Würth ferrite (I LOVE the actual bass, more and better than without it), so I do not modify the woofer filter, which is key to flattening the response.

 

[napilopez]

LS50W is a DSP-based system so wouldn't flattening it be very small R&D cost and zero per-unit cost?

I'm not a speaker designer so someone correct me if I'm wrong, but DSP only gets you so far if the dip is a directivity issue with the actual hardware. It clearly appears to be a directivity issue as a midrange dips appear on the on-axis curves for the LS50, LS50W, R3, R5, R7, and R11 but dissappear at outer angles. See: R series whitepaper. The dips are less apparent in KEFs measurements in part due to the 'squished' scaling, but they are there.

You can't really address such dips with simple EQ in the DSP, as removing it in the on-axis response will introduce a peak in the off-axis response. I don't know much about active crossovers but I'm not sure you'd be able to fix such a directivity issue in that realm either.

My impression is KEF prioritizes the early reflections and power response curves over the on-axis and listening window curves, as the former two probably best reflect the timbral balance we'll hear in a typical mid or farfield listening setup. Notably, they are consistently the smoothest curves in measurements of KEF speakers.

In any case, it's also worth remembering that the coaxial design has a huge advantage in the vertical response. These relatively small dips should ultimately have far less of an effect on the power response than the huge crossover dips that show up in the vertical plane with almost every traditional driver array.

 

[Soniclife]

So to put it in a little perspective, KEF was a British company for 31 years, then a British Colonial company for approx. 5 years, and "becoming" Chinese due to UK/China treaties/agreements over rule of Hong Kong itself. Gold Peak has owned KEF for nearly half of its existence. The factory in China is wholly owned by KEF/GP Acoustics, and staffed by their employees. This is not a case of a company "farming out" manufacturing to whichever factory meets the bid.

All Reference Series loudspeakers have always been manufactured in England, as have "concept" speakers and flagships (Maidstone aka Reference 109, Concept Blade, Blade, Blade 2, Muon). Some of their best work has been produced in the Gold Peak era. I think this makes them unique among old respected brands which face financial difficulties and are acquired by others, which typically become a shadow of their glory, and then a name only, slapped onto pure junk. Sadly the fate of Nakamichi, once one of the best audio brands. Countless others as well.

Those who are interested in companies that have a respect for research, engineering, testing, and value, should look into the history of KEF. Materials science (not just choosing a material because it is expensive or exotic lol), anechoic testing of FR and directivity, computer modeling, sharing specifications and test data with customers, matching drivers, etc. have all been hallmarks of KEF since practically the beginning.

Yes, you can see from my signature that I'm a long-term KEF customer. I have even more of their speakers that aren't listed in my sig haha! Everything I've bought from them over the years is still going strong and is fully original. The first loudspeaker I ever heard that could begin to approximate the voice of my favorite singer was the KEF Reference 107, in 1987. I've only been disappointed in two products I've bought from them over the years, one was the tiny Picoforte 1 speaker with iPod dock (lol, I shouldn't have expected a miracle for something so small, even from KEF haha, but my sister has the set now and loves it), and the other is the M400 headphone, which sounds great but has a design deficiency (IMO) that can lead to issues with the wiring to one of the drivers over time. If I ever need to replace my current speakers, I hope KEF will offer me a loyalty discount lmao!

Edited to add: I hope the OP will forgive my off-topic post, but the subject of KEF's ownership and history kept popping up, so I thought it might be worthwhile to give a bit of information for those who might be interested.

Is all the R&D still done in the UK?

 

[Hugo9000]

Is all the R&D still done in the UK?

I think most of the key people are based there, but I would assume they also have work being done in Hong Kong and at their mainland China facility, because it just seems logical to me that they would use whatever top staff they have in all locations doing valuable work. I don't blame them for keeping important work in England, or for emphasizing that connection. It has "good optics" with regard to the idea of a continuous history of the company, while allowing them to market their more expensive lines as being part of a British tradition of handcrafted luxury goods.

Here is a Stereophile article about a visit to the UK facilities:
https://www.stereophile.com/content/three-days-england-kef

And another recent account of a UK factory tour I just found:
https://www.13thnote.net/2019/09/02/kef-a-factory-tour/

The only article I've found about anyone touring their Asian facilities is this one by Darko:
https://darko.audio/2016/08/kef-gp-acoustics-no-longer-your-fathers-loudspeaker-company/

According to that, all "acoustic" research is still conducted in the UK, "electronics" research is conducted in HK and their mainland China factory, and "product development" takes place at the Hong Kong headquarters.