KEF LS50 Bookshelf Speaker Review

[MZKM]

No. They are very much alike due to concentric driver. Look at the fine detail and you see the difference.

Uhh, it showed horizontal for both, here is the link to the one that was duplicate:
https://www.audiosciencereview.com/...izontal-contour-audio-measurements-png.47778/
Change 47778 to 47777 and you get the first horizontal plot, change to 47779 and you get the reflection plots.

BTW, sequential numbering is poor for security, I can change that to any number, say "12345" and pull up whatever attachment that is; not so much that important here, but Tom Scott mentions it in this video:

 

[oldsysop]

Another toy speaker, and expensive

 

[amirm]

Sorry if I missed this somewhere but I've been meaning to ask does the Klippel take into account port output in any way or are these graphs purely from the woofer? Thanks.

It doesn't care what makes up a speaker. It evaluates as radiating sound waves all around the speaker. You could produce that with one driver or 100. With or without port. It doesn't care. It computes and solves the wave equations based on sampled measurements (520 in this case).

 

[beefkabob]

They didn't sound good to me either.

 

[TimW]

This narrowly beats the JBL 305P in preference rating which matches my experience with the two speakers.

 

[Cahudson42]

FWIW) I don't expect that comparison to be pretty -- that's not a reason not to do it, but... manage one's expectations.

Ah sure - but then there is always $100 DSP - the miniDSP IL-DSP to 'pretty it up'

Or even maybe better, applying the 2x4 HD with native B652 AIR cap/resistor 'crossover' removed and the woofer and AMT DSP optimized directly..

I will also posit that 'size matters' and a 6.5" $22 woofer properly high cut-off just might give interesting results compared to a 5.25" $700 one... 27 sq in vs. 42 is not insignificant.

 

[amirm]

FYI and this is a serious note: I tested the LS50 twice. One without accuracy with respect to acoustic center, etc. and one with. I actually realigned the system again. It made no difference in measurements other than the polar plot at the end.

 

[napilopez]

FYI and this is a serious note: I tested the LS50 twice. One without accuracy with respect to acoustic center, etc. and one with. I actually realigned the system again. It made no difference in measurements other than the polar plot at the end.

Good to hear! Something I'm not clear about, by the way: Does the Klippel let you "shift" the '0' point in post? (No not trying to get into the reference axis debacle here again, just curious!)

How was the polar plot different?

Either way, good to see that consistency.

 

[amirm]

How was the polar plot different?

The axis in 2-d Polar plot pointed up a few degrees at 20 kHz from what I recall.

This is not easy business. You have to align a microphone and poke it into the tweeter without going into the tweeter. One wrong move on axis control and instead of down, you have be going left and you go right into the driver! So I can only risk it so much for getting the microphone positioned right.

Then there is the business of pointing a speaker at 90 degrees to the microphone. With the odd shapes of speakers, there is no good way to do this. I try to center them but I am not accurate to one or two degrees, that's for sure.

 

[Spocko]

Another toy speaker, and expensive

Pretty much it's interior design friendly with many color to choose from.

 

[napilopez]

The axis in 2-d Polar plot pointed up a few degrees at 20 kHz from what I recall.

This is not easy business. You have to align a microphone and poke it into the tweeter without going into the tweeter. One wrong move on axis control and instead of down, you have be going left and you go right into the driver! So I can only risk it so much for getting the microphone positioned right.

Then there is the business of pointing a speaker at 90 degrees to the microphone. With the odd shapes of speakers, there is no good way to do this. I try to center them but I am not accurate to one or two degrees, that's for sure.

I appreciate the insight! When I had the KEF R3s I accidentally dinged the midwoofer from setting it up too close for neafield measurement, leading to one of the biggest cringes of my life. Didn't affect the sound but bothered me to no end.

 

[hardisj]

Distortion Measurements
These are all in-room measurements at the same SPL as the spinorama measurements:

Here is the same data but presented in the more familiar THD ratio/percentage:

Amir, exactly how is this test conducted? "The same SPL as the Spinorama"... so, 2.83v/1m? Or is this calculated based on the nearfield measurements? The 5% THD stood out like a sore thumb to me but when I went and looked at my test results of the *RAW LS50 WOOFER DRIVER* I had about 5% THD below 100hz so from that we know the culprit is indeed the driver itself.

Here's my testing of the raw driver (woofer and tweeter) pulled from wayback.
https://web.archive.org/web/20180316184816/http://medleysmusings.com/kef-ls50-drive-unit

And here's pictures of my testing of the woofer for distortion at 90dB @ 1m & 96dB @ 1m equivalent.

 

[NTK]

Amir, exactly how is this test conducted? "The same SPL as the Spinorama"... so, 2.83v/1m? Or is this calculated based on the nearfield measurements? The 5% THD stood out like a sore thumb to me but when I went and looked at my test results of the *RAW LS50 WOOFER DRIVER* I had about 5% THD below 100hz so from that we know the enclosure isn't making things worse.

Here's my testing of the raw driver (woofer and tweeter) pulled from wayback.
https://web.archive.org/web/20180316184816/http://medleysmusings.com/kef-ls50-drive-unit

And here's pictures of my testing of the woofer for distortion at 90dB @ 1m & 96dB @ 1m equivalent.

View attachment 47803View attachment 47801View attachment 47802

Amir explained it in an earlier post. The spinorama test is probably not the ideal test for HD, since the input signal is pretty low (2.83 Vrms for passives).

 

[mhardy6647]

FYI and this is a serious note: I tested the LS50 twice. One without accuracy with respect to acoustic center, etc. and one with. I actually realigned the system again. It made no difference in measurements other than the polar plot at the end.

My simple-minded assessment of this observation is that it speaks well of the design and execution of the system -- or am I underthinking?

Ah sure - but then there is always $100 DSP - the miniDSP IL-DSP to 'pretty it up'

Or even maybe better, applying the 2x4 HD with native B652 AIR cap/resistor 'crossover' removed and the woofer and AMT DSP optimized directly..

I will also posit that 'size matters' and a 6.5" $22 woofer properly high cut-off just might give interesting results compared to a 5.25" $700 one... 27 sq in vs. 42 is not insignificant.

Actually, now that you mention it -- that is an interesting idea (especially with full-blown DSP not just for EQ, but a deliberately implemented XO).
I suspect that the skeevy enclosures of the B652-AIR would thwart 'high end' performance, though. I mean, yeah, that could potentially be ameliorated, but one quickly gets to the Stone Soup (or Nail Broth)* school of loudspeaker modification -- wherein, basically, there's nothing left of the original loudspeaker. This approach, in total seriousness, is often applied to the "Heritage" Klipsch loudspeakers.

The "AIR's" woofer's pretty nominal, too, irrespective of its size (methinks).

_____________
* You all know one or the other of these basically interchangeable fables, yes?
If not, Danny Kaye's version of the latter tale (which I grew up with) is marvelous.

Sorry for the digression.

 

[EB1000]

Thanks for the review. before I got my Focal Aria 906, I've auditioned both the Aria 906 and the LS50 at home. I've run Dirac live calibration. The Aria 906 sounded way better to me, with better bass response and authority. Only imaging was slightly better on the LS50, and only when I sat closer to them. I've decided to go with the Aria 906, which I got for a very good deal (800$ a pair for the black gloss finish).


I would love to see ASR review of the Aria 906.

 

[TimW]

My simple-minded assessment of this observation is that it speaks well of the design and execution of the system -- or am I underthinking?


Actually, now that you mention it -- that is an interesting idea (especially with full-blown DSP not just for EQ, but a deliberately implemented XO).
I suspect that the skeevy enclosures of the B652-AIR would thwart 'high end' performance, though. I mean, yeah, that could potentially be ameliorated, but one quickly gets to the Stone Soup (or Nail Broth)* school of loudspeaker modification -- wherein, basically, there's nothing left of the original loudspeaker. This approach, in total seriousness, is often applied to the "Heritage" Klipsch loudspeakers.

The "AIR's" woofer's pretty nominal, too, irrespective of its size (methinks).

Sorry for the digression.

That AMT tweeter should probably be high-passed in the 3-5 KHz range. That basic 6 1/2" Poly cone woofer should probably be low-passed in the 1.5-2.5 KHz range, at least if you want to get good directivity behavior from it. There is bound to be a severe directivity mismatch between these two drivers and no amount of DSP or enclosure modification can fix that.

 

[Ron Texas]

I have a pair of these. With a sub, they are awesome.

 

[hardisj]

Amir explained it in an earlier post. The spinorama test is probably not the ideal test for HD, since the input signal is pretty low (2.83 Vrms for passives).

Well, the input signal is OK. But ideally you would want to conduct additional tests at higher output, I agree.

I was more curious at what point in space the Klippel derived the distortion measurements from. It doesn't look like it's as simple as "nearfield", it seems.

The perplexing thing is that the FR in the THD measurement results should mimic the on-axis FR. In the case of what Amir posted it does not. It looks more like the sound power (even though it obviously is not). You can see what I mean in this example (scroll down about halfway).
https://web.archive.org/web/20180209082123/http://medleysmusings.com/dyn-audio-esotar²-430-midrange/

Before anyone gets their panties in a bunch ... I AM NOT SAYING THE DATA IS WRONG. I am simply asking how it is measured. If we don't know that's fine. Just say so. I can email my contacts at Klippel directly and ask if necessary.

 

[milosz]

To me all box speakers sound like.... boxes. I'll take my Quad ESL-57 speakers or my tri-amped Magneplanar MG 3.6's over ANY monkey coffins. Open baffle GR research / Rhymik servo subs complete the boxless system.

 

[Audio Monkey]

Great review! I like the continued evolution of the measurements! The LS50 is a popular speaker and one I considered recently for a desktop setup given the price drop to $900/pair. Although generally very favourable, some reviews mention a hint a brightness, and the peak around 5khz supports those comments. Looks like the speaker would benefit from being placed near a wall and/or perhaps room correction/equalization.