Finally a Bluetooth speaker worth buying!
Coming soon to a listening room near you (Free shipping and free returns (at least here in Australia))
Buzzword compliant - AirPlay, BT, wifi streaming via Chromecast, Spotify and Roon ready etc
Teeny tiny skinny floorstanding cabinets your born in 21C partner / children would like.
for the DIY audio nerds- it’s an
active 3 way with a 5” wide baffle with 4” coaxial midrange/tweeter closely surrounded by four 5.25” woofers crammed into only 12.5L of cabinet volume.
500/100/100 Watts (4ohm) for LF/MF/HF
Uses DSP for baffle step compensation, delays, crossover, equalisation and +/- phase correction, and limiters for woofer protection.
The coaxial is part of KEF’s recently released ranges, albeit smaller, with acoustically tuned damper for the tweeter.
Each pair of midwoofers share a common magnet (!) , but optimised for low rubber surround radiation distortion: the white paper suggests a very flat and symmetrical Kms(x) & BL(x) curves.
The woofers also use voice coil current negative feedback for reduced bass distortion, which is advantageous when using small woofers in small boxes.
Here is the White Paper with the details:
https://us.kef.com/pub/media/ls60-content/LS60W_WP.pdf
Now, I know what you’re thinking. Big deal.
And they’re €7K/US$7K/AU$10K a pair.
Let's forget about the price for a moment, or whether you can afford them, and even if you could, are you in the market for one.
I'm glad that not all wireless speakers are $100 BT speakers... I'm looking at you Apple, Google with more resources for R&D than anybody else. This is a problem that Sean Olive commented on- how do you get people to be interested in acoustic excellence; there’s not much captured in the specs or reviews that indicate why a large high performance speaker (the size and weight of a person and the cost of a small automobile) is better than a $100 speaker with essentially the same frequency response? Of course; that is, without having to sell them a piece of (audio) furniture?
Come on, let’s make an easy-to-use wireless speaker that the kids and spouse can use with their voice / phone, but us audio nuts might also enjoy listening to...
K E F
Make Speakers Great Again
Coming soon to a listening room near you (Free shipping and free returns (at least here in Australia))
Buzzword compliant - AirPlay, BT, wifi streaming via Chromecast, Spotify and Roon ready etc
Teeny tiny skinny floorstanding cabinets your born in 21C partner / children would like.
for the DIY audio nerds- it’s an
active 3 way with a 5” wide baffle with 4” coaxial midrange/tweeter closely surrounded by four 5.25” woofers crammed into only 12.5L of cabinet volume.
500/100/100 Watts (4ohm) for LF/MF/HF
Uses DSP for baffle step compensation, delays, crossover, equalisation and +/- phase correction, and limiters for woofer protection.
The coaxial is part of KEF’s recently released ranges, albeit smaller, with acoustically tuned damper for the tweeter.
Each pair of midwoofers share a common magnet (!) , but optimised for low rubber surround radiation distortion: the white paper suggests a very flat and symmetrical Kms(x) & BL(x) curves.
The woofers also use voice coil current negative feedback for reduced bass distortion, which is advantageous when using small woofers in small boxes.
Here is the White Paper with the details:
https://us.kef.com/pub/media/ls60-content/LS60W_WP.pdf
Now, I know what you’re thinking. Big deal.
And they’re €7K/US$7K/AU$10K a pair.
Let's forget about the price for a moment, or whether you can afford them, and even if you could, are you in the market for one.
I'm glad that not all wireless speakers are $100 BT speakers... I'm looking at you Apple, Google with more resources for R&D than anybody else. This is a problem that Sean Olive commented on- how do you get people to be interested in acoustic excellence; there’s not much captured in the specs or reviews that indicate why a large high performance speaker (the size and weight of a person and the cost of a small automobile) is better than a $100 speaker with essentially the same frequency response? Of course; that is, without having to sell them a piece of (audio) furniture?
Come on, let’s make an easy-to-use wireless speaker that the kids and spouse can use with their voice / phone, but us audio nuts might also enjoy listening to...
K E F
Make Speakers Great Again
Last edited:
https://audioxpress.com/news/kef-un...l-in-one-high-fidelity-streaming-audio-system
A cross between LS50 wireless and the Blade. Interesting that Navin sketched a very similar box just days ago in a discussion in the New Mark Audio Drivers thread.
Yes, if money was no concern, and i had a use for them they are very interesting.
dave
So make one....much cheaper, and satisfying.
Hint: party speakers have most of those functions, and have a rechargeable battery as well.
They start here about US$ 30, and up to about $300 for LG and Panasonic class...I have not kept in touch with the latest prices.
Hint: party speakers have most of those functions, and have a rechargeable battery as well.
They start here about US$ 30, and up to about $300 for LG and Panasonic class...I have not kept in touch with the latest prices.
I have built one. 4 x Peerless 830070 in an ML-V with FF85wKeN as midTweeter in MidTL. Going to swap an Alpair 5.2/3 into them. Not rectangualr thou. A shape that will probably never be built again. Now veneered in solid Yew.
dave
dave
Last edited:
What do you mean Dave? It looks great! As in, you won't repeat that?
The white paper states that KEF has an internal volume of just 12.7L for four 5.25" woofers. That's the way to go with active- put it in a small box and EQ for desired bass extension (within reasonable/target limits)
A shootout between the new LS60, Buchardt A700 and Dynaudio Focus 60 XD would be interesting, all price comparable and active speakers with built in wireless functions. Throw in a pair of LS50W2 standing on a pair of BK Double Gems as a wild card at half the price and have fun.
Worth noting the cone area of the four woofers is equivalent to a 10” but obviously with less Xmax.
Now for DIY, four of the new Dayton Epique 5.5” subwoofers and an Alpair of your choice or 10f/8424 in WAW configuration for the mid/highs or even a Seas 6” coaxial (maybe based on the Seas KingRoy swapping out the woofer) could be interesting powered by Hypex plate amps. Not wireless but add a couple of WiiM mini and off you go.
Worth noting the cone area of the four woofers is equivalent to a 10” but obviously with less Xmax.
Now for DIY, four of the new Dayton Epique 5.5” subwoofers and an Alpair of your choice or 10f/8424 in WAW configuration for the mid/highs or even a Seas 6” coaxial (maybe based on the Seas KingRoy swapping out the woofer) could be interesting powered by Hypex plate amps. Not wireless but add a couple of WiiM mini and off you go.
Why is "new product introduction" threads becoming more and more a part of this site - why/how is this DIY? Do you perhaps work for KEF?
//
Because as an (amateur) speaker designer I don't work in a vacuum.?
Do you make drivers yourself? Do you make amps yourself? Do you make capacitors or inductors yourself?
Do you wind your own voice coils?
Can't we be interested in new products lest being accused of working for the company producing the product?
Here's my own bi-cabinet design, high sensitivity, wide dispersion 3 way I made whilst waiting for Purifi woofers to be made.
Dual sealed 25cm woofers, 6" midrange, 1" soft-dome tweeter, all from different manufacturers.
Phase coherent, optimised for listening distance 3+ metres, large domestic rooms or recording studios
System sensitivity 92dB@2.83V, 112+ * dB SPL @1m capable per speaker.
Olive Preference Score 7+
In room F3 of 30 Hz, without any EQ boost. with selectable boost easily down to 20Hz.
*likely more, but yet to be verified...
If I bought the KEF LS60, would I hang up my boots and call it a day?
Nah.. we love DIY audio too much.
Do you make drivers yourself? Do you make amps yourself? Do you make capacitors or inductors yourself?
Do you wind your own voice coils?
Can't we be interested in new products lest being accused of working for the company producing the product?
Here's my own bi-cabinet design, high sensitivity, wide dispersion 3 way I made whilst waiting for Purifi woofers to be made.
Dual sealed 25cm woofers, 6" midrange, 1" soft-dome tweeter, all from different manufacturers.
Phase coherent, optimised for listening distance 3+ metres, large domestic rooms or recording studios
System sensitivity 92dB@2.83V, 112+ * dB SPL @1m capable per speaker.
Olive Preference Score 7+
In room F3 of 30 Hz, without any EQ boost. with selectable boost easily down to 20Hz.
*likely more, but yet to be verified...
If I bought the KEF LS60, would I hang up my boots and call it a day?
Nah.. we love DIY audio too much.
Attachments
Last edited:
It is REALY hard to build. Lots of angles.
From the way they desribe it they have just taken two of the compound push-push woofers from their little sub into the rectangle.
dave
I wouldn’t count on that. Given the development and the purpose of these woofer si would be surprisded if they did not have more tarvel than a typical 10” woofer.
dave
The
This one is 5.25" . You probably haven't had a chance to look at the whitepaper in detail from but BL(x) and KMS(x) charts the Klippel limited excursion is about 6mm excursion. What's just as important is the symmetry of the BL(x), KMS(x) and Le(x).
A typical 5.25" driver has 85cm^2, so 4 would have about 300cm^2; about as much as a typical 10"
Even so, 4 x 5.25" in 12.5L is a nice achievement, and dynamic range compression for sub-bass is a wise move for longevity.
KC62 is dual 6.5"
This one is 5.25" . You probably haven't had a chance to look at the whitepaper in detail from but BL(x) and KMS(x) charts the Klippel limited excursion is about 6mm excursion. What's just as important is the symmetry of the BL(x), KMS(x) and Le(x).
A typical 5.25" driver has 85cm^2, so 4 would have about 300cm^2; about as much as a typical 10"
Even so, 4 x 5.25" in 12.5L is a nice achievement, and dynamic range compression for sub-bass is a wise move for longevity.
Last edited:
Hi, I noticed you have quoted a preference rating for your speakers, I have not been able to find out what I should aim for on this, can I assume from your comment that anything above 7 is a good point to head for (my current design is at 6.75 at the moment with a -0.67dB/octave tilt on room response). Any guidance will be gratefully received.
Single number doesn't tell much about how speaker sounds. But to answer your questions checkout audio science review, top numbers are 9+ I think.
It is possible to make speaker just for high number to sell more while for example the transient response could be poo and sound boring even though the CTA graphs look great. The number does not include distortion, or info about dynamics, SPL capability, suitability of directivity / size / cost to your situation etc.
Certainly, if the number is great there is high possibility the sound is great, or at least the designer is capable of making great sounding speaker even if the number was rigged by cost of sound quality. On the other hand lower (any) number speaker could potentially sound better in your application if it is just more suitable for your application. Speaker scoring 10 could fart out at party and obviously that is not good party speaker regardless of best number in the universe.
This is not to defend low numbers or high numbers or any numbers, just that people would be aware what the number presents and what it doesn't.
The KEF looks like very good speaker, very high tech, I've got no other concerns it wouldn't have very high performance and sound top than perhaps the SPL capability, its kind of enough I suppose but not sure if very good dynamics due to being that small after all. Directivity is very well in control by the graphs in the white paper but perhaps some could prefer narrower pattern. All in all, very much a problem free speaker to my eye with great features. If I had mo money I'd buy it to be a benchmark for diy builds 😀
It is possible to make speaker just for high number to sell more while for example the transient response could be poo and sound boring even though the CTA graphs look great. The number does not include distortion, or info about dynamics, SPL capability, suitability of directivity / size / cost to your situation etc.
Certainly, if the number is great there is high possibility the sound is great, or at least the designer is capable of making great sounding speaker even if the number was rigged by cost of sound quality. On the other hand lower (any) number speaker could potentially sound better in your application if it is just more suitable for your application. Speaker scoring 10 could fart out at party and obviously that is not good party speaker regardless of best number in the universe.
This is not to defend low numbers or high numbers or any numbers, just that people would be aware what the number presents and what it doesn't.
The KEF looks like very good speaker, very high tech, I've got no other concerns it wouldn't have very high performance and sound top than perhaps the SPL capability, its kind of enough I suppose but not sure if very good dynamics due to being that small after all. Directivity is very well in control by the graphs in the white paper but perhaps some could prefer narrower pattern. All in all, very much a problem free speaker to my eye with great features. If I had mo money I'd buy it to be a benchmark for diy builds 😀
Last edited:
@tmuikku many thanks for the pointers, I am building a sealed stand mount (old style monkey box) with WO24P-4, EM1308 and ET448 and Nanodigi DSP crossover. LR2 slopes model well with good phase coherence on all three drivers and it is flat on axis, so looking good so far, next job is driver cut outs then I can start measuring but will use sim as a starting point only. Sorry to go off thread, back to the LS60, intermission over.
Ugg10,
The preference rating is based on research at least a decade or two old, and it has some limitations.
As of 13/05/22 the highest ranked speaker with an publicly available spinorama, currently in production is the new KEF Blade 1 Meta, with a score of 7.7:
https://pierreaubert.github.io/spinorama/scores.html
But I've heard the KEF Blade 1 (non-Meta) and Muon and got turned off by the space age looks. Sorry, looks matter.
The so called top scores that tmuikku on ASR is referring to, is with a theoretical "perfect subwoofer" integrated and further equalised following the use of Klippel NFS. No actual speaker in the original study scored more than 7.x And gook luck integrating a subwoofer perfectly into your system, but I guess we all have to strive for something.
Anyway, the decimal point means nothing: they are not statistically significant. This from Sean Olive himself.
Secondly, @hifijim might be able to contribute here, there's a lot of extrapolation going on.
BUT, in terms of conventional box speakers it's goal it is, at least, sometime to aim for even tonal/spectral balance and neutrality, which is a Harman speaker characteristic. Like a 7.x speaker is more neutral than a 5.x speaker, all things else considered equal, and more likely to be preferred by the listener.
Aiming for smooth flat on-axis only is going to lead you more astray than the complete spinorama. Then again, the Preference score favours the smooth and flat predicted in room response. Well, again, that's predicted. Your actual in-room response is going to be dependant on your room. And THAT's the difficult part, the final frontier IMHO, which is what John Krevskovsky and the late Sigfried Linkwitz was working on- how to be free of the room's influences. Or Syng, how to auto-detect and measure the room and calibrate to any random room.
From VituixCAD2 Re: Preference score:
https://kimmosaunisto.net/Software/VituixCAD/VituixCAD_help_20.html#Preference_rating
[paste]
Predicted preference rating is calculated according patent application US 2005/0195982 A1 by Sean Olive when equation 9 or 10 is selected with the following exceptions:
Equation provides also mathematical approach with less illogicalities and conflicts to be valid for optimizing:
Second, please do your crossovers with actual measurements, simulations only get you so far. You should not, and cannot, reliably simulate if you haven't taken complete polar measurements prior to simulating. Kimmo has expressed his distaste for people using sims like there's no tomorrow. It really only goes so far, and making very accurate precise measurements in a skill/art in itself. Kind of like speaker cabinet building.
Then verify your actual design with full spinorama measurements + distortion measurements + phase measurements + listening measurements and go back to crossover.
My design goals are high sensitivity, high dynamic range, (concert level), crossovers in phase, wide dispersion, impulse response, acoustics crossovers in phase. Smooth on axis, power response is a given, and a lot of designers have been working this way for decades eg. Dennis Murphy, Rick Craig, Jeff Baby.
Everyone one ASR wants to know the Preference Score, I also used that pane to hide my crossover. 😉
The preference rating is based on research at least a decade or two old, and it has some limitations.
As of 13/05/22 the highest ranked speaker with an publicly available spinorama, currently in production is the new KEF Blade 1 Meta, with a score of 7.7:
https://pierreaubert.github.io/spinorama/scores.html
But I've heard the KEF Blade 1 (non-Meta) and Muon and got turned off by the space age looks. Sorry, looks matter.
The so called top scores that tmuikku on ASR is referring to, is with a theoretical "perfect subwoofer" integrated and further equalised following the use of Klippel NFS. No actual speaker in the original study scored more than 7.x And gook luck integrating a subwoofer perfectly into your system, but I guess we all have to strive for something.
Anyway, the decimal point means nothing: they are not statistically significant. This from Sean Olive himself.
Secondly, @hifijim might be able to contribute here, there's a lot of extrapolation going on.
BUT, in terms of conventional box speakers it's goal it is, at least, sometime to aim for even tonal/spectral balance and neutrality, which is a Harman speaker characteristic. Like a 7.x speaker is more neutral than a 5.x speaker, all things else considered equal, and more likely to be preferred by the listener.
Aiming for smooth flat on-axis only is going to lead you more astray than the complete spinorama. Then again, the Preference score favours the smooth and flat predicted in room response. Well, again, that's predicted. Your actual in-room response is going to be dependant on your room. And THAT's the difficult part, the final frontier IMHO, which is what John Krevskovsky and the late Sigfried Linkwitz was working on- how to be free of the room's influences. Or Syng, how to auto-detect and measure the room and calibrate to any random room.
From VituixCAD2 Re: Preference score:
https://kimmosaunisto.net/Software/VituixCAD/VituixCAD_help_20.html#Preference_rating
[paste]
Predicted preference rating is calculated according patent application US 2005/0195982 A1 by Sean Olive when equation 9 or 10 is selected with the following exceptions:
- Responses are smoothed 1/33 oct (instead of 1/20 oct bands) for AAD and NBD to avoid excessive smoothing due to 1/48 oct step of internal frequency points.
- LFX=1.162 and LFQ=0.5 when 'with sub' is checked. This equals to flat response with Butterworth 18 dB/oct HP at 17.5 Hz and limits rating to 10.0 with equation 9.
- NBD is sensitive to slope; rating gets worse if response slope is not 0.
- SM is sensitive to slope; rating gets worse if absolute value of response slope is not 1.
- LFX and LFQ fail with directive LF radiators because sound power at LF is lower than LW reference at low mid.
Equation provides also mathematical approach with less illogicalities and conflicts to be valid for optimizing:
- NBD is calculated with moving average to eliminate effect of slope.
- SM is calculated as r^2 with slope normalized to -1.0 to eliminate effect of slope.
- LFX is LFX_SP or LFX_LW which one is smaller to support also directive LF radiators having SP significantly below reference at low mid (y_LW).
- LFQ is LFQ_SP or LFQ_LW which one is smaller.
- Optional SL_ON and SL_LW with adjustable slope targets are available. SL_ON and SL_LW are for optimizing with custom equation to avoid unstability if optimized crossover parameters have strong effect to slop
Second, please do your crossovers with actual measurements, simulations only get you so far. You should not, and cannot, reliably simulate if you haven't taken complete polar measurements prior to simulating. Kimmo has expressed his distaste for people using sims like there's no tomorrow. It really only goes so far, and making very accurate precise measurements in a skill/art in itself. Kind of like speaker cabinet building.
Then verify your actual design with full spinorama measurements + distortion measurements + phase measurements + listening measurements and go back to crossover.
My design goals are high sensitivity, high dynamic range, (concert level), crossovers in phase, wide dispersion, impulse response, acoustics crossovers in phase. Smooth on axis, power response is a given, and a lot of designers have been working this way for decades eg. Dennis Murphy, Rick Craig, Jeff Baby.
Everyone one ASR wants to know the Preference Score, I also used that pane to hide my crossover. 😉
Last edited:
Oh and another thing:: Let's remind our-selves that the range of 20dB is the difference between 1W and 100W!
When designing loudspeakers high performance loudspeakers I like to work on a 20dB scale, or less.
Did anyone catch why 50dB scale was settled on as the CTA2034A standard?
Here's my speaker again in the usual 50dB scale:
When designing loudspeakers high performance loudspeakers I like to work on a 20dB scale, or less.
Did anyone catch why 50dB scale was settled on as the CTA2034A standard?
Here's my speaker again in the usual 50dB scale:
^ Everything looks better zoomed out, even earth looks peaceful place from outer space. 🙂
tktran303, if you fancy bigger number for the rating you could add slants/big roundovers for the top box and / or waveguide the tweeter. Basically minimize empty flat baffle. SPL capability and everything would be about the same, except the graphs would be little smoother due to reduced diffraction. Perhaps some adjustment for crossover. This is key for smooth graphs, smooth measurements, "void of acoustic issues".
tktran303, if you fancy bigger number for the rating you could add slants/big roundovers for the top box and / or waveguide the tweeter. Basically minimize empty flat baffle. SPL capability and everything would be about the same, except the graphs would be little smoother due to reduced diffraction. Perhaps some adjustment for crossover. This is key for smooth graphs, smooth measurements, "void of acoustic issues".
Yes. Of course. Well aware of improvements can be had with large round-overs and facets,
done decades ago by the likes of Peak Consult, Rockport etc
but these cabinets are prototypes made by my semi-retired cabinet maker; mitre joins and thus square edges only.
I had to work with what I had (leftover drivers, leftover amps, leftover DSP etc)- it’s a one off design, not for mass/commercial production, whilst awaiting Purifi woofers and tweeters.
done decades ago by the likes of Peak Consult, Rockport etc
but these cabinets are prototypes made by my semi-retired cabinet maker; mitre joins and thus square edges only.
I had to work with what I had (leftover drivers, leftover amps, leftover DSP etc)- it’s a one off design, not for mass/commercial production, whilst awaiting Purifi woofers and tweeters.
Last edited:
I have built loudspeakers using 2 x 4, 2 x 5.25 and 4 x 4 for bass. Not subwoofer depth, or at ear bleeding levels but do a good job down to about 35 Hz. I expect with DSP these will be more capable, althou perhaps limited by the small volume.
I expect they will sell many.
dave