I paired mine with a E15HP sealed sub and I found 100 -110 Hz seemed to work very well.
-
Welcome to ASR. 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
- Thread starter amirm
- Start date
I paired mine with a E15HP sealed sub and I found 100 -110 Hz seemed to work very well.
Smooth as possible for studio like use based amirm's sample, it cost 5 dB sensitivity, listening axis horisontal shall be 20º offset else the usual 4,5kHz bump will grow few dB for Reference angle curve, can share EQ filter curve as wav-file for convolution engine should anyone like a listen.
dc655321
Major Contributor
- Joined
- Mar 4, 2018
- Messages
- 1,601
- Likes
- 2,279
Not sure what the probe signal is here, nor what the sampling rate and FFT size are. But, this is how one may discern what would be necessary to achieve the desired resolution.
According to this document, the NFS uses the TRF module. Here's the AN for TRF:
https://www.klippel.de/fileadmin/klippel/Bilder/Our_Products/R-D_System/PDF/S7_TRF.pdf
They’re also at accesories4less refurbished or recertified.
@dc655321 and @hardisj Thanks for the info. My impression was that zero padding (in time domain) is interpolation in the frequency domain. It doesn't add information. From the Klippel AN for TRF, the time scale of the example CSD plot is totally different (in hundreds of ms) than Amir's plot (10 ms max). That's why I am puzzled.
I don't know for sure but that confusion may be from an apples to oranges comparison. 10ms is only really useful for high frequency CSD plots; seeing ringing in the upper midrange or tweeter. When I saw your comment that they were showing hundreds of milliseconds I assumed they were measuring decay time of a room. The AN has this little bit under the CSD example: "Box resonance and other unwanted modes are so detected easily."
Which implies to me they were measuring enclosure resonance in the example graphic they gave. And the fact the measurement is so low in frequency backs that up; you'd need at least 50ms to get 20hz accuracy and the additional time seems like it would be suited for a low frequency CSD.
This, though, is just all another reason why I sometimes feel that CSD are shown for the sake of it. Time windowing is important and how you adjust the window can drastically alter the result. If I can correlate CSD to FR then it's just extra info to back up what the FR shows. But I don't find it terribly useful on its own.
To be honest, they're not that bad, but still overpriced and totally not suited to nearfield usage, with an on-axis like that. I'm sure their very nice look have convinced more people than their performance.
Wonder how they compare to active solutions around their price like the KSD C5-Reference or the much cheaper Tannoy Gold 5/7 and /Fluid Audio FX50.
Wonder how they compare to active solutions around their price like the KSD C5-Reference or the much cheaper Tannoy Gold 5/7 and /Fluid Audio FX50.
Listening tests are just a small part of these reviews, but I suppose you could argue that for subjective listening impressions it would be more fair to compare against another speaker in the same category.
On the other hand, it's a pretty good argument for going active. And even among passives, there are better choices for bass extension - the buchardt s400 is smaller than the LS50, for instance, while having much deeper bass (albeit it's also more expensive). You could argue the LS50 should be crossed with a sub, but that's not inherent to their design.
In today's market, the LS50's best quality is that they can be had for pretty cheap and are very EQable. Plus they're gorgeous =] But the R3 and LS50W appear to be much better choices if you don't want to do extensive eq, have got the cash, and want that Uni-Q goodness.
By the way, it's interesting that over the past few years, most of the arguments about LS50 passive vs LS50W have centered around the amplification and DSP magic. Lots of people (including myself at some point when I was dumber, I'm pretty sure) have claimed that the LS50W is better than the usual LS50 for something roughly like "KEF knows its hadware best, so it can match the amps perfectly to the Uni-Q's woofer and tweeter!" but that the LS50 can maybe live up to the LS50W if you have a super powerful fancy $10,000 amp.
But by now, it's clear the answer for why the LS50W sounds better is the most obvious one: it has the better frequency response. Unfortunately there isn't much in the way of high-resolution data for both speakers, especially from the same source, but even then it's pretty clear KEF "tidied up" the LS50W's response. From Hifinews:
Forum member @Erik had previously posted this comparison from hifi-test.de reviews:
And here's my own capture of the LS50W (only out to 60 degrees, and no bass, but shows a smoother on-axis albeit with more of a tilt).
Most notable is that the low-mids scoop is cleaned up, and the upper mids through treble are significantly better balanced. There's still a scoop around 1.6Khz, but it's a lot more innocuous as its now the biggest anomaly in the frequency response and cleans up in the off-axis response. So KEF probably sacrificed a bit of DI smoothness for a cleaner PIR. The tilt can also probably be accounted for a bit with the app if needed.
Once again, frequency response is king.
But by now, it's clear the answer for why the LS50W sounds better is the most obvious one: it has the better frequency response. Unfortunately there isn't much in the way of high-resolution data for both speakers, especially from the same source, but even then it's pretty clear KEF "tidied up" the LS50W's response. From Hifinews:
Forum member @Erik had previously posted this comparison from hifi-test.de reviews:
And here's my own capture of the LS50W (only out to 60 degrees, and no bass, but shows a smoother on-axis albeit with more of a tilt).
Most notable is that the low-mids scoop is cleaned up, and the upper mids through treble are significantly better balanced. There's still a scoop around 1.6Khz, but it's a lot more innocuous as its now the biggest anomaly in the frequency response and cleans up in the off-axis response. So KEF probably sacrificed a bit of DI smoothness for a cleaner PIR. The tilt can also probably be accounted for a bit with the app if needed.
Once again, frequency response is king.
Near-field is more forgiving of on-axis, as any slight head movement would bring you off-axis, the listening window will be more influential compared to far-field.
However, the question remains whether the tonal balance (~1dB/oct lope down, with maybe a bass boost) is still the preferred when in near-field.
Ron Texas
Master Contributor
- Joined
- Jun 10, 2018
- Messages
- 6,655
- Likes
- 10,421
In my setup right now the sub starts to roll off at 50Hz. Measurements indicate the slope of the single sub's low pass is probably 3rd order. The LS50's are rolled off 6db @ 50 HZ using the 4th order high pass filter built into my Crown XLS 1502.
The crossover point appears to be at 75 hz. I would rather take it higher, but that produces directional bass. It might be time to break down and buy a pair of L12's and get the benefit of lower distortion, increased dynamic range. The sweet spot with one sub is good enough now.
edit:Rythmik L12's are back ordered, probably for month.
The crossover point appears to be at 75 hz. I would rather take it higher, but that produces directional bass. It might be time to break down and buy a pair of L12's and get the benefit of lower distortion, increased dynamic range. The sweet spot with one sub is good enough now.
edit:Rythmik L12's are back ordered, probably for month.
Last edited:
While that's right, that bad on-axis is more dominant than the more decent reflections in nearfield. I can't even call them good room filling speakers, as they probably don't like SPL.
I really like that curved front, though.
confucius_zero
Addicted to Fun and Learning
- Joined
- Nov 22, 2018
- Messages
- 542
- Likes
- 351
Wait... so a LSR305 is better than ls50?!
Better for the money (for far-field and if you don't have hum issues).
But no, the LS50 scores slightly better.
Hmm my impression has always been the opposite, that in the nearfield I place much more priority on the direct sound. Or perhaps more accurately, I think of a smaller window: narrower horizontally with more relative weighting to the vertical. I also suppose it also depends on whether you're discussing the listening window with respect to toe-in vs listening position.
My caution with the spinorama listening window is that it was derived from devantier and that it was clearly aimed at living room setups. The closest listening position of any of the rooms evaluated was just about two meters, with the average being over 3 meters (10'10") - which doesn't describe the majority of what people colloquially call nearfield listening. I wonder if there's any research on what the typical listening window is for studios. I think the NRC listening window of 15 degree horizontal 15 degree vertical is probably more accurate, maybe even smaller. I'm personally much more open to vertical movements than horizontal when listening in the nearfield - big horizontal shifts mean big changes to stereo imaging.
In any case, I'd still consider these a fair bit worse for listening nearfield than farfield. PIR looks a lot better to me than the listening window.
confucius_zero
Addicted to Fun and Learning
- Joined
- Nov 22, 2018
- Messages
- 542
- Likes
- 351
amazing compilation there! more representative than the pink panthers. Could there be one on the other product types?
Similar threads
- Poll
