• 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 LS50 Bookshelf Speaker Review

D700

Senior Member
Joined
Dec 21, 2018
Messages
311
Likes
370
I've retired my LS50s to TV sound duty, running them with a Devialet 120. With the Devialet's SAM feature, the KEFs were unlistenable on movies: the bass was simply overloading anything else the speakers could do.

To resolve the problem, I picked up a $399 SVS SB-12 NSD from Amazon. Because I've been too busy and too tired to do measurements in the family room, I quickly found a recommendation from KEF that subwoofers be crossed over to the LS50s at 79 Hz and followed it, setting a 4th-order crossover to the KEFs at 79 Hz in the Devialet.

The sound is now superb for TV use, and with most listeners off-axis anyway, I think the LS50s plus a subwoofer are more or less ideal for two-channel "home theater." I will test it with music in due course.
Yeah, I followed same recommendations and use them with Velodyne Microvee, 50/50 Music/TV in a den, very happy.
 

maty

Major Contributor
Joined
Dec 12, 2017
Messages
4,596
Likes
3,160
Location
Tarragona (Spain)
I went from the Q150 to the LS50. My experience was that the LS50's cabinet is better built. Measurements suggest it is more innert as far as resonances go. That may or may not make it worth the doubling of price. As I understand it any ringing or resonances are a time domain problem and as a result cannot be effectively EQ'd out.

I certainly like the look of the blacked out LS50.

I was expecting more low frequency extension from the LS50 but that is not the case, they roll off very similarly and I was unsatisfied with the setup (given the cost) until I added a subwoofer (F12-300).

I read somewhere that the Q350 might be the better buy overall for more extension, and while I disregarded that advice initially they might have been on to something.
.

To listen to music, only the coaxial 5.25" has the magic. Either they have improved the Uni-Q 6.5" or I would avoid it. Of course, with very good recordings and high DR.

I am sure that with more mass + damping the Q150 would sound much better, as mine Q100. And with better components at the crossover. I have done the first process on some loudspeakers for years and always noticed improvement. With the Q100 the sound improvement was substantial!
 
Last edited:

maty

Major Contributor
Joined
Dec 12, 2017
Messages
4,596
Likes
3,160
Location
Tarragona (Spain)
With the years they have been in the market there is enough information on Internet about how they are to be placed, and not just the KEF manual.
 

AudioJester

Addicted to Fun and Learning
Forum Donor
Joined
Jan 7, 2020
Messages
912
Likes
1,211
Interesting discussion. I have both Rythmik and SVS subs, and prefer the servo sub by a long margin. Obviously many factors involved.
Howevee, can the Klippel system measure subs?
 

MZKM

Major Contributor
Forum Donor
Joined
Dec 1, 2018
Messages
4,240
Likes
11,462
Location
Land O’ Lakes, FL
Howevee, can the Klippel system measure subs?
I believe Amir said it can.

Side note, @amirm, can the Klippel state at what simulated distance the measurements are being taken at after doing the extrapolation? Meaning, can it tell us how far away one needs to be for the drivers to fully integrate? I think that would be an interesting aspect to note.
 

MZKM

Major Contributor
Forum Donor
Joined
Dec 1, 2018
Messages
4,240
Likes
11,462
Location
Land O’ Lakes, FL
Yes. But since response of a sub varies a ton in a real room, I am not sure of the value.
CEA-2010 rating for max passing SPL. As I mentioned elsewhere, data-bass.com is the best site for subwoofer measurements, Rythmik F18 as an example.

This is a max SPL discussion you can't get out of :p.

Group delay is also good (i guess waterfall and spectrogram show the same), it shows if the bass is "tight" or "loose".
 

AudioJester

Addicted to Fun and Learning
Forum Donor
Joined
Jan 7, 2020
Messages
912
Likes
1,211
Yes. But since response of a sub varies a ton in a real room, I am not sure of the value.
Amir, would there me an advantage to having servo control for a subwoofer that would show up in measurements? Or is it all in my mind?
 
OP
amirm

amirm

Founder/Admin
Staff Member
CFO (Chief Fun Officer)
Joined
Feb 13, 2016
Messages
44,368
Likes
234,386
Location
Seattle Area
Side note, @amirm, can the Klippel state at what simulated distance the measurements are being taken at after doing the extrapolation? Meaning, can it tell us how far away one needs to be for the drivers to fully integrate?
It can and it does. I don't currently have confidence in what it is showing however. As I test larger speakers, I can better tell how good that metric is. Here is what it looks like for KEF LS50 for example:
1580419161101.png


The highest order function flattens at about 0.6 meter where I have drawn that arrow. That means from there on we are in far field. Someone less lazy than me can do the math and tell us if this is correct:

1580419457855.png
 
OP
amirm

amirm

Founder/Admin
Staff Member
CFO (Chief Fun Officer)
Joined
Feb 13, 2016
Messages
44,368
Likes
234,386
Location
Seattle Area
Amir, would there me an advantage to having servo control for a subwoofer that would show up in measurements? Or is it all in my mind?
It should show in distortion measurements assuming the implementation is good. The feedback mechanism varies between different implementations, some measuring current, others having sensors or secondary windings on the woofer itself.
 

MZKM

Major Contributor
Forum Donor
Joined
Dec 1, 2018
Messages
4,240
Likes
11,462
Location
Land O’ Lakes, FL
It can and it does. I don't currently have confidence in what it is showing however. As I test larger speakers, I can better tell how good that metric is. Here is what it looks like for KEF LS50 for example:
View attachment 47994

The highest order function flattens at about 0.6 meter where I have drawn that arrow. That means from there on we are in far field. Someone less lazy than me can do the math and tell us if this is correct:

View attachment 47996
Hmm, for this wouldn't you need to choose the frequency of the crossover region? Meaning, if it's mostly just the woofer playing 1000Hz, I don't know if the far-field distance for that is what's needed to answer the question.
 

jamestown

Member
Forum Donor
Joined
Jan 17, 2020
Messages
5
Likes
3

Attachments

  • in room response, no eq no sub.jpg
    in room response, no eq no sub.jpg
    66.8 KB · Views: 218
OP
amirm

amirm

Founder/Admin
Staff Member
CFO (Chief Fun Officer)
Joined
Feb 13, 2016
Messages
44,368
Likes
234,386
Location
Seattle Area
CEA-2010 rating for max passing SPL. As I mentioned elsewhere, data-bass.com is the best site for subwoofer measurements, Rythmik F18 as an example.

This is a max SPL discussion you can't get out of :p.
I am aware of that and hence the reason I am not sure frequency response measurement is of much value relative to how loud it can play.

That said, I need to build a larger platform for the Klippel system as subs can be quite a bit wider than speakers. I have bought an aluminum plate but then need to find the time to machine it and drill holes to mount it.
 
OP
amirm

amirm

Founder/Admin
Staff Member
CFO (Chief Fun Officer)
Joined
Feb 13, 2016
Messages
44,368
Likes
234,386
Location
Seattle Area
Hmm, for this wouldn't you need to choose the frequency of the crossover region? Meaning, if it's mostly just the woofer playing 1000Hz, I don't know if the far-field distance for that is what's needed to answer the question.
You mean the graph I posted or the text? The graph gives that answer implicitly. The more complex the soundfield, the higher the order needed to describe it. But once there, you have it regardless of what the source looks like.
 

bobbooo

Major Contributor
Joined
Aug 30, 2019
Messages
1,479
Likes
2,079
I am aware of that and hence the reason I am not sure frequency response measurement is of much value relative to how loud it can play.

That said, I need to build a larger platform for the Klippel system as subs can be quite a bit wider than speakers. I have bought an aluminum plate but then need to find the time to machine it and drill holes to mount it.

The LFQ variable in Sean Olive's preference rating paper, defined as "Absolute average deviation (dB) in bass response from LFX to 300 Hz" in the sound power curve relative to the mean sensitivity in the listening window, had a contribution close to 20% in predicting preference in his initial formula. Admittedly this was based only on the first set of measurements (which I think were mostly/all stand-mounted speakers), but I think this is still good evidence that low-frequency quality (i.e. bass flatness), not just bass quantity (max SPL) or extension (LFX), does have a significant correlation with preference, and so LFQ should definitely be calculated for subwoofers (requiring the NFS to measure their frequency response), and possibly somehow incorporated into a subwoofer preference score.
 
Last edited:

BYRTT

Addicted to Fun and Learning
Forum Donor
Joined
Nov 2, 2018
Messages
956
Likes
2,452
Location
Denmark (Jutland)
I went from the Q150 to the LS50. My experience was that the LS50's cabinet is better built. Measurements suggest it is more innert as far as resonances go. That may or may not make it worth the doubling of price. As I understand it any ringing or resonances are a time domain problem and as a result cannot be effectively EQ'd out.....

If those resonances measured is 100% reflection free minimum phase (MP) EQ also called analog or IIR filter in digital domain will smooth out those resonances both in amplitude domain and time domain, you can see it in examle below that MP response curves always have a phase pattern (time) that belongs to that particular response curve so on computers one curve is enough to compute/extract the other.

Cant know but will imagine when you talk the ringing cant be effectively EQ'd out as in below examples its because measured curve is not 100% reflection free so one try EQ out some none MP response and therefor waterfall wont clean up as it should, for acoustic domain measurements other reason to get errors than not 100% reflection free sweeps could be we often use few ADC/DAC units into measurement chain at varius rates and some of them can add their portion of excess phase if not calibrated.
MP.png
 
Last edited:

aarons915

Addicted to Fun and Learning
Forum Donor
Joined
Oct 20, 2019
Messages
685
Likes
1,140
Location
Chicago, IL
This Spin is fairly close to other measurements of the LS50 but one oddity I notice is the clear bump at 5k that I haven't seen in other measurements. Most seem to show an elevated region from 2-5k on axis but as you go off-axis, the 5K region smooths out while the 2k Resonance remains. The 2k region is almost always stronger than 5k except in this spin, which is a bit odd. Room measurements show the same thing, I've noticed it in my spatially averaged measurements and a few in this thread have posted measurements showing the 2k area with more energy than the 5k. I don't really have a question in all of that, just wanted to mention it since no one has brought it up.
 
Top Bottom