I swapped Ls50 for an R3 meta ….im I imagining

[312elements]

Is it your perspective that having the woofer adds to the diffraction by disrupting the baffle face below the concentric driver? I envision this could be the case. Perhaps that is why on the Blades and LS60 KEF moved the woofers to the sides, but I don't know.

My perspective is that all speakers are a compromise.

The R3, like most speakers keeps the drivers close enough together so that the sound seems to be emanating from one place. It's a very well designed speaker. The LS50 does that trick better by not having a second woofer at all. Rather than being close enough not be problematic, it's an absolute best case scenario. In the same way that 3 ways have some advantages over 2 ways, a point source driver (or coaxial) has advantages over a dual alignment. The extra woofer isn't causing diffraction issues. The signal visually looks different emanating from the speaker.

The diffraction in regard to enclosure design takes that one step further. Again, you've got a best case scenario here for imaging.

With regard to the LS60 and the Blade, my hypothesis is that it has more to do with WAF and internal resonances. You could use various materials to dampen a resonance as is often done, the dual opposed woofers on those designs cancel the resonances and stop them before they become a problem that has to be solved. It's pretty cool tech, especially the way KEF does it with the shared magnet design. It's actually one unit with two drivers that spans across the enclosure.

The R3 does lots of things better than the LS50. There's more excursion in the coaxial driver in the LS50 and that excursion negatively impacts other aspects of its performance. If this, along with the deeper bass appeals to you, then it's a better candidate.

As a big fan of full range speakers I've built many variations borrowing ideas from KEF along the way and testing theories. Full range speakers have their own strong points and weaknesses and for many the compromises aren't worth it, but it's hard to find a speaker that images better than a well made full range driver in a well designed enclosure. The bass will likely be lackluster, if the driver is too large, you will have beaming issues. They typically perform best without toe-in (because the frequency response is most even 15 degrees off axis) which can cause problems with reflected response but they image like nobodies business. All compromises.

 

[MAB]

You reckon ? Is the ones like a ls50 on steroid ? What would be the benefit .. remember I play not too loud .. I would say 75 at 3 metres is the loudest .. I routinely listen at less admittedly peak taken from a phone … but I’m seeing 60-64 at night
Where would I benefit ? How ?

I think you are getting people's personal preferences, it's inevitable in a forum. The Genelec's aren't necessarily about loud, they just sound good, but so do KEF. Swapping out a pair of Genelecs for either of these very good KEF are not likely going to magically fix your perception of imaging problems.

 

[Tangband]

My perspective is that all speakers are a compromise.

The R3, like most speakers keeps the drivers close enough together so that the sound seems to be emanating from one place. It's a very well designed speaker. The LS50 does that trick better by not having a second woofer at all. Rather than being close enough not be problematic, it's an absolute best case scenario. In the same way that 3 ways have some advantages over 2 ways, a point source driver (or coaxial) has advantages over a dual alignment. The extra woofer isn't causing diffraction issues. The signal visually looks different emanating from the speaker.

The diffraction in regard to enclosure design takes that one step further. Again, you've got a best case scenario here for imaging.

With regard to the LS60 and the Blade, my hypothesis is that it has more to do with WAF and internal resonances. You could use various materials to dampen a resonance as is often done, the dual opposed woofers on those designs cancel the resonances and stop them before they become a problem that has to be solved. It's pretty cool tech, especially the way KEF does it with the shared magnet design. It's actually one unit with two drivers that spans across the enclosure.

The R3 does lots of things better than the LS50. There's more excursion in the coaxial driver in the LS50 and that excursion negatively impacts other aspects of its performance. If this, along with the deeper bass appeals to you, then it's a better candidate.

As a big fan of full range speakers I've built many variations borrowing ideas from KEF along the way and testing theories. Full range speakers have their own strong points and weaknesses and for many the compromises aren't worth it, but it's hard to find a speaker that images better than a well made full range driver in a well designed enclosure. The bass will likely be lackluster, if the driver is too large, you will have beaming issues. They typically perform best without toe-in (because the frequency response is most even 15 degrees off axis) which can cause problems with reflected response but they image like nobodies business. All compromises.

Well spoken !

 

[meracus]

I think you are getting people's personal preferences, it's inevitable in a forum. The Genelec's aren't necessarily about loud, they just sound good, but so do KEF. Swapping out a pair of Genelecs for either of these very good KEF are not likely going to magically fix your perception of imaging problems.

Of course … and this is just a rabbit hole thingy …. The thinking .. large waveguide concentric down to the bass driver , infinitely tweakable … cabinet design with what appears to be anti diffraction ( by way of shape ) … it strikes a chord , you can’t help to see the similarities and of course the gains … if anything , a call to a dealer / purchase or return …. You know , to ascertain …
I did my fair shameful share of gear swapping in the last 3 decades … maybe let’s have an educated episode !
I really like the Genelec idea too … plus of course gonna try the placement thingy lower stand whachmacallit , plus I have 30 days return policy … worse case I get back to the 50 … the goal was to improve upon it

 

[MAB]

A crossover at 420 Hz is audible if its not perfect . And a passive crossover never is.
Also the drivers are not better than +-1 dB in similarity and in the r3 meta there are three of them .

You often bring up driver and crossover component mismatch as being a source of all types of sonic differences.
I have tried to show in the past what to expect with component mismatches. It's a surprisingly small effect, especially when put in an enclosure with crossover. I won't go search for my many responses with data to you, but perhaps I need to start a driver matching thread to try to disentangle these speculations from reality.

In the mean time, do you have a specific detail to share, or is it speculation that there is >1dB mismatch, and that this mismatch is relevant? Are you suggesting OP do a driver matching exercise? And, it looks like the raw driver matching is much more accurate than you claim, any actual data to share on 0.5dB matching. It would be nice to see data from you on actual driver mismatch, and actual measured effect. It's quite difficult when so many problems are attributed to this effect, and even more so when the actual mismatch is much less than you claim.

The R3 and LS50 have different in-room response. The sound at MLP after in-room reflections are going to be different, and may need different placement to obtain the same spatial effects. Or different treatment for the same placement. These are worth many dB, not 1dB or less.

 

[MAB]

Of course … and this is just a rabbit hole thingy …. The thinking .. large waveguide concentric down to the bass driver , infinitely tweakable … cabinet design with what appears to be anti diffraction ( by way of shape ) … it strikes a chord , you can’t help to see the similarities and of course the gains … if anything , a call to a dealer / purchase or return …. You know , to ascertain …
I did my fair shameful share of gear swapping in the last 3 decades … maybe let’s have an educated episode !
I really like the Genelec idea too … plus of course gonna try the placement thingy lower stand whachmacallit , plus I have 30 days return policy … worse case I get back to the 50 … the goal was to improve upon it

Yes, placement. Looking at Erin's reviews of the two, the speakers will interact in your room differently. Different ratios of in-room reflections will alter image. I have no idea what you in room measurements are, but likely even these two speakers from the same manufacturer will have variations larger than a few dB! To try to illustrate, let me share a slightly different version of room interactions. I recently rearranged my room treatment. I replaced underperforming acoustic wall panels behind my mains by moving bookcases behind my speakers. Here is my setup after moving the bookcases.

I measured the in-room response with the old crappy panels, and with the bookcases. I was able make both measurements without moving the speakers, only the panels were replaced with the bookcases. The mic and furniture placement did not change, and I did not change DSP settings on the speakers between measurements. Please ignore the speaker brand and model!

Oh my, order of magnitude change in early vs. late sound ratio in the right speaker! Looking at my room, it makes sense. Staring at the individual traces, many of the reflection changes are more than 5dB, even some 10dB changes!

I can assure you, imaging dramatically improved with the bookcases-as-treatment. After re-running GLM's DSP correction, I was able to get much better in-room response compared to previous since GLM had many less problems to deal with, but that is a sperate exercise. The point is after moving the bookcases, I have more flexibility and larger sweet-spot, no surprise given the measurements. Humorously, my analog vinyl collection is having a positive and measurable impact on my sound!

To my point, simple changes in room (placement, furniture, treatment) are responsible for large changes worth many dB in reflected sound and in-room response. The converse, small differences in your two KEF options will likely need different placement and room setup for imaging, especially if you have large amounts of reflected sound.

 

[terryforsythe]

The extra woofer isn't causing diffraction issues. The signal visually looks different emanating from the speaker.

On the KEF R3 Meta the published specification for the crossover frequency is 420 Hz. The wavelength at that frequency is 81.7 cm. The distance between the center of the woofer and the center of the coaxial driver looks to be somewhere around somewhere around 20 cm (based on looking at KEF's picture). Unless there is an issue with the crossover, at 420 Hz, the 20 cm spacing probably is not audible on axis from the listening position.

I have been tuning my Elac Uni-Fi reference bookshelf speakers, which have a coaxial midrange/tweeter and a woofer. The passive crossovers are removed. I have experimented with woofer/midrange crossover frequencies between 300 Hz and 500 Hz. If the offset of the woofer from the coaxial driver was an issue from an imaging perspective, moving the crossover frequency lower should have improved the imaging. I did not notice that happening, though.

Where I have detected an impact on sound stage and imaging is tuning the midrange/tweeter crossover frequency and addressing cabinet resonances. Specifically, the crossover frequency with the best optimized directivity index (the best of which the speakers are capable) has resulted in the largest sound stage and best imaging. Also, dampening cabinet resonances by adding glue fillets to the cabinet joints, adding an additional internal brace, and adding a lot of fiberglass insulation (ports are plugged) helped improve sound stage and imaging. The Elacs have a curved baffle, and that does help reduce diffraction.

The R3 meta has a directivity index quite a bit better than my Elacs, so I doubt that is its issue. However, the R3 metas have a flat baffle and 90 degree corners, which will result in worse diffraction than the baffle shape of the LS50. Also, the LS50 meta has slightly better dispersion. I think these are where the difference in imaging performance arise. The LS50s also may have less cabinet resonance issues, but I don't know. If so, this also may help their imaging.

 

[Tangband]

You often bring up driver and crossover component mismatch as being a source of all types of sonic differences.
I have tried to show in the past what to expect with component mismatches. It's a surprisingly small effect, especially when put in an enclosure with crossover. I won't go search for my many responses with data to you, but perhaps I need to start a driver matching thread to try to disentangle these speculations from reality.

In the mean time, do you have a specific detail to share, or is it speculation that there is >1dB mismatch, and that this mismatch is relevant? Are you suggesting OP do a driver matching exercise? And, it looks like the raw driver matching is much more accurate than you claim, any actual data to share on 0.5dB matching. It would be nice to see data from you on actual driver mismatch, and actual measured effect. It's quite difficult when so many problems are attributed to this effect, and even more so when the actual mismatch is much less than you claim.

The R3 and LS50 have different in-room response. The sound at MLP after in-room reflections are going to be different, and may need different placement to obtain the same spatial effects. Or different treatment for the same placement. These are worth many dB, not 1dB or less.

There are examples where the matching between similar drivers are very good. However, I can easily show two examples where they are not :

www.audioexcite.com » SEAS H1456 ER18RNX

www.audioexcite.com » SB Acoustics “Satori” MW13P-8

Seas er18rnx

Sample 1 = Blue
Sample 2 = Red

Satori MW13P

1 dB is not much, but its audible .

If the Kefs are better than this, its ofcourse a good thing.

The main reason for the single source speaker to bring a better spatial illusion for the listener is probably because the sound is coming from a single point and theres no phase issues from nonperfect executed crossovers.

 

[Vacceo]

On the KEF R3 Meta the published specification for the crossover frequency is 420 Hz. The wavelength at that frequency is 81.7 cm. The distance between the center of the woofer and the center of the coaxial driver looks to be somewhere around somewhere around 20 cm (based on looking at KEF's picture). Unless there is an issue with the crossover, at 420 Hz, the 20 cm spacing probably is not audible on axis from the listening position.

I have been tuning my Elac Uni-Fi reference bookshelf speakers, which have a coaxial midrange/tweeter and a woofer. The passive crossovers are removed. I have experimented with woofer/midrange crossover frequencies between 300 Hz and 500 Hz. If the offset of the woofer from the coaxial driver was an issue from an imaging perspective, moving the crossover frequency lower should have improved the imaging. I did not notice that happening, though.

Where I have detected an impact on sound stage and imaging is tuning the midrange/tweeter crossover frequency and addressing cabinet resonances. Specifically, the crossover frequency with the best optimized directivity index (the best of which the speakers are capable) has resulted in the largest sound stage and best imaging. Also, damping cabinet resonances by adding glue fillets to the cabinet joints, adding an additional internal brace, and adding a lot of fiberglass insulation (ports are plugged) helped improve sound stage and imaging. The Elacs have a curved baffle, and that does help reduce diffraction.

The R3 meta has a directivity index quite a bit better than my Elacs, so I doubt that is its issue. However, the R3 metas have a flat baffle and 90 degree corners, which will result in worse diffraction than the baffle shape of the LS50. Also, the LS50 meta has slightly better dispersion. I think these are where the difference in imaging performance arise. The LS50s also may have less cabinet resonance issues, but I don't know. If so, this also may help their imaging.

I wonder, if the shape of the LS50 is a clear advantage, why hasn´t it copied on the R series? Price and manufacturing complexities?

 

[terryforsythe]

I wonder, if the shape of the LS50 is a clear advantage, why hasn´t it copied on the R series? Price and manufacturing complexities?

Only KEF can answer that question. My guess is the aesthetics they wanted for the R series.

Nonetheless, look at the shape of the Blades. KEF clearly spent much effort to reduce diffraction on those.

Also, on the LS60 KEF used a curved baffle with a large radius on the corners, which reduce diffraction and probably were intended to do so. Indeed, looking straight on at the LS60 it is difficult to see that the baffle is curved. Only when looking straight down on it from the top is it really noticeable. So, I doubt the curved baffle was used on the LS60 simply for aesthetic reasons. I think that diffraction control played a part in the design process.

 

[thewas]

Only KEF can answer that question. My guess is the aesthetics they wanted for the R series.

Also it is a cost factor, the LS50 exists only as a compact model while for the R series they would have to make curved baffles and enclosures also for the floorstanders which they even avoided at their Reference series.

 

[Ricardojoa]

Why dont you try to reposition them ? They are different speakers and should be positioned up accordingly.

 

[Mnyb]

Hmm is it not so that the shadowflare around the coax is there to control the diffraction issues of the square box so it's not as bad as it looks ?
The same shadow flare known to not be properly in place , some user had to show into place with some force ? .

The LS60 solves the issue (allegedly Kef says.. ) by having the baffle so very narrow that the problem is less pronounced given the xover between tweeter and mid , it's almost directly at the driver and the tweet only "sees" the midrange cone and flare acoustically .

The blade is much bigger and thus needs the roundover , and actually is it not so that a small rounding only 1-2cm is useless ? it must be a big roundover like the blade or the whole cab in LS50 and LSX ? otherwise you migth use a square box .

I remember my old Thiel's I once had it had large roundings of the whole baffle surface .

Try some new positions height's and toe in .

Or if its very small listening space why not the active LS50 and subwoofers there you go .

 

[Tangband]

Some of these responses are just absurd. Like really really absurd. The combination of the point source speaker with the heavily radiused baffle are the main reasons that LS50 images so well. You're not hearing things. Your brain is not playing tricks on you. It has zero to do with the frequency response or any of the other other nonsense being touted and you're not wrong for preferring it. This isn't complicated.

I would also go as far as to say that a really good fullrange driver does certain things in an even better way than a good coaxial driver - noticable in the soundstage department .

The technical reason might be explained by Erin :

Kali Audio IN-5 3-Way Studio Monitor Review

Kali Audio IN-5 3-Way Studio Monitor Review

www.erinsaudiocorner.com

”As you can see in the below graphic, the position of the cone does influence the high frequency response. Depending on frequency, the effect is as much as 2dB.”

A true fullrange single driver dont have this specific modulation problem , because the whole driver is moving.
In a coaxial, the tweeter is fixed.

 

[Sokel]

I would like to echo the possibility that speaker height affects things more than you think. I remember comparing some speakers once and I thought I heard a major change in imaging quality that turned out to have more to do with height. Both speakers sounded worse when at the height of a some furniture in the room. As soon as I elevated them a few inches to above the furniture, the sound cleared up. Playing white noise made the sound difference even clearer at different heights. It was subtle, but consistently noticeable. It was like I had diffusers or absorbers, except it was a shelf full or records.

Another very subtle thing about the LS50 vs R3 (meta or not), is that there is a smoother roll-off past 90 degrees, as @thewas mentioned. Whether this really has an effect for every listening space ehhh idk, but it might have an effect in some spaces. I did think the LS50W and LS50 Meta imaged slightly better than the R3, and I've heard some people echo similar thoughts. I've also heard people voice the opposite, but usually the former. Could just be a sighted bias thing with the full point source vibe, but still.

On that note though, you might want to try angling the R3s a bit different from the metas too.

At the end of the day measurements are a super useful guide, but there are always individual room factors and preferences that influence things.

Totally agreed.
This is how I can turn mine from a living nightmare to me (elevated presence) to decent (measurement at about a meter but it partly follows the pattern even 3m away at MLP):



and with higher resolution,legends describe the height)

 

[thewas]

”As you can see in the below graphic, the position of the cone does influence the high frequency response. Depending on frequency, the effect is as much as 2dB.”

Please keep in mind that is +-1 mm excursion with a 9V battery which is rather a "proof of concept" than a realistic level, when such a mid driver makes 100+ dB usually it has other more significant problems.

A true fullrange single driver dont have this specific modulation problem , because the whole driver is moving.

A true fullrange single driver suffers from much worse doppler modulation due to whole cone movement, which together with the usually not optimal directivity pattern are their biggest disadvantages and the reasons pretty much nobody uses them for serious neutral reproduction like monitoring.

 

[Tangband]

Please keep in mind that is +-1 mm excursion with a 9V battery which is rather a "proof of concept" than a realistic level, when such a mid driver makes 100+ dB usually it has other more significant problems.


A true fullrange single driver suffers from much worse doppler modulation due to whole cone movement, which together with the usually not optimal directivity pattern are their biggest disadvantages and the reasons pretty much nobody uses them for serious neutral reproduction like monitoring.

Please show me measurements that is proof of that !

 

[thewas]

Please show me measurements that is proof of that !

Here you go, it basic physics in the end, the higher the driver displacement and wider the frequency band the larger such FM/Doppel distortions are:
https://www.silbersand.de/Doppler.html
Same happens by the way also to multitone/intermodulation distortions where adding more ways significantly reduces them (unfortunately the Neumann paper about it of their old site isn't online anymore):
https://www.audiosciencereview.com/...speakers-unimportant.11659/page-4#post-336502

 

[thewas]

unfortunately the Neumann paper about it of their old site isn't online anymore

Thankfully it exists still here, the IMD part starts at page 15:

https://audiophilestyle.com/s3/article-images/2020/0713/Monitors-Glossary-measurement-descriptions_EN.pdf

 

[Tangband]

Thankfully it exists still here, the IMD part starts at page 15:

https://audiophilestyle.com/s3/article-images/2020/0713/Monitors-Glossary-measurement-descriptions_EN.pdf

Thanks for the interesting links . Now - is there a possibility that newer loudspeaker drivers might behave better ?
This is the IMD from a Markaudio chr120 loudspeaker which shows excellent results.

Markaudio CHR120 Test and Review

Note: These plans can be purchased here. In this blog post I test and review the Markaudio CHR120 fullrange driver. This is a 12cm driver featuring a magnesium cone and dust cap. This driver is slightly larger than a typical 6.5" driver, with and SD of 147 cm^2 which is about the same as the...

josephcrowe.com

Intermodulation Distortion​

”Playing a multitone test signal (12 band per octave) at 85dB test signal (1m) produces the following IMD. Distortion is at -60dB (2kHz) in the midrange and -67dB in the upper treble (10kHz)”

”Increasing the test SPL to 95dB is revealed below. Midrange and treble distortion reduces to -50dB. Bass distortion at 100Hz remains low at -50dB. This is an excellent result especially considering the driver's size.”