D'Appolito Style Coaxial vs Single Woofer (R2C vs R3)

[aarons915]

You can ask the same question then how would 2way coax beat the actual point source one Full Range driver in a box that doesn't have to fake it, of course FR is more limited in spl and has higher distortion.

FR doesn't controll anything, 2way coax doesn't controll the directivity in midrange, and 3way doesn't controll the directivity in bass.

Correct, a 2-way Coax and a full range driver would be nearly equal in that regard but a full range driver has too many compromises compared to a coax in my opinion. Not sure what the rest of your post has to do with anything, I was just saying a 3 way that is crossed over at 500Hz isn't a better point source than the 2-ways even with woofers flanking the UniQ driver.

 

[73hadd]

This is a really compelling idea. I really like the LS50 and wondered why even though the measurements are better on R3, the reputation and use of the LS50 has held up really well, and continue to be used by reviewers as a reference. It's possible that the proliferation is better on the LS50 and maybe not as many have heard the R3. I have also heard anecdotes that the LS50 is preferred over the R3 among some who have heard both.

All this is to say, I invented in my mind the idea that the "point source" was the key to the LS50 magic. And a cabinet that feels like a cement block.

Using the R2c would certainly regain the symmetry. I like this idea a lot. I am not sure if these additional points will add or subract:

-LS50 vocals do not sound "in a box" to me
-R7 vocals DO sound "in a box" to me, and the overall tone not quite as good. The additional cabinet and porting, maybe taking away from "point source?"
-Reference 5 - this speaker completely disappears. It sounded a shade more mellow than neutral to me.
-R3 - I've never listened to it
-R2c - sealed cabinet! I would like to hear them as a stereo pair!
-R2c - I would like to see Amir measure them. I don't think he will like them though.

On that last point, I am not sure KEF does the "play loud" role as well, for reasons that have been discussed before (moving waveguide?). However if you are listening in stereo I think the KEF coax design brings a better level of imaging at lower sound levels.

 

[aarons915]

This is a really compelling idea. I really like the LS50 and wondered why even though the measurements are better on R3, the reputation and use of the LS50 has held up really well, and continue to be used by reviewers as a reference. It's possible that the proliferation is better on the LS50 and maybe not as many have heard the R3. I have also heard anecdotes that the LS50 is preferred over the R3 among some who have heard both.

All this is to say, I invented in my mind the idea that the "point source" was the key to the LS50 magic. And a cabinet that feels like a cement block.

Using the R2c would certainly regain the symmetry. I like this idea a lot. I am not sure if these additional points will add or subract:

-LS50 vocals do not sound "in a box" to me
-R7 vocals DO sound "in a box" to me, and the overall tone not quite as good. The additional cabinet and porting, maybe taking away from "point source?"
-Reference 5 - this speaker completely disappears. It sounded a shade more mellow than neutral to me.
-R3 - I've never listened to it
-R2c - sealed cabinet! I would like to hear them as a stereo pair!
-R2c - I would like to see Amir measure them. I don't think he will like them though.

On that last point, I am not sure KEF does the "play loud" role as well, for reasons that have been discussed before (moving waveguide?). However if you are listening in stereo I think the KEF coax design brings a better level of imaging at lower sound levels.

I would say the R3 is a better speaker in stock form but the LS50 has a lot of potential if you can EQ them and set them up with subs and a slightly higher crossover point so they can play loud cleanly.

I generally agree with your other assessments, the R series really makes me wonder some things about speaker design in general and whether it's a good idea to cross a speaker over in the middle of the vocal range and/or fundamental frequency range in general. I think this could be why many people seem to like the coherency of a 2 way because a single driver is playing all of the fundamentals and into the harmonics and doesn't sound weird when those frequencies are handed off to a tweeter above 2k or so. Splitting the frequencies between 2 drivers at 400hz sounds a bit odd because we're not used to hearing vocals for example, coming from 2 different places. This is why I think the best option is for a 3 way to cross as low as possible but definitely under 300Hz or turn a 2 way into a 3 way with subs and a slightly higher crossover point.

 

[MrPeabody]

Larger and more expensive.

The problem of using R2c is that their frequency response plummets below 100 Hz, unlike the larger towers.

View attachment 121750

Certainly if you compare the R2C to the large towers there will be a world of difference in bass response. But compared to the R3, in theory the R2C will have a less abrupt rolloff even though the F3 point is evidently higher (per Kef's specifications). So I'm not sure that I would say that the frequency response of the R2C "plummets" below 100 Hz. The graph you included looks like it is synthetically generated in some fashion below 200 Hz, but even this graph indicates a slow rolloff of 12 dB per octave, which is a curiously close match to the theoretical rolloff for sealed speakers.

 

[richard12511]

I was just saying a 3 way that is crossed over at 500Hz isn't a better point source than the 2-ways even with woofers flanking the UniQ driver.

That is true if you think it's a good thing that all of the sound is coming from the same place. I think some people actually prefer the bigger presentation of the sound coming from a taller source.

In the latest Triangle Espirit Antal Tower speaker review, @amirm compared the Revel M16 against the Triangle. He admits that the tonality was better with the Revel, but that he still preferred the Triangle because the sound was emitted from a taller vertical space. With the Revel, the sound sounded too much like a point source, which was enough to make him prefer a tonally much worse speaker. The LS50 would do even worse than the M16 did in that comparison.

May be one of those personal preference things.

 

[Matias]

Certainly if you compare the R2C to the large towers there will be a world of difference in bass response. But compared to the R3, in theory the R2C will have a less abrupt rolloff even though the F3 point is evidently higher (per Kef's specifications). So I'm not sure that I would say that the frequency response of the R2C "plummets" below 100 Hz. The graph you included looks like it is synthetically generated in some fashion below 200 Hz, but even this graph indicates a slow rolloff of 12 dB per octave, which is a curiously close match to the theoretical rolloff for sealed speakers.

See the difference for yourself.

R3

R2c

 

[MrPeabody]

See the difference for yourself.

R3
View attachment 122018


R2c
View attachment 122019

See what for myself? Is there something here that objectively validates the use of the word "plummet" to describe the slow 12 dB per octave rolloff of the R2C? What word would you then have in reserve for describing the much steeper rolloff of the R3 below its port tuning frequency?

 

[Matias]

See what for myself? Is there something here that objectively validates the use of the word "plummet" to describe the slow 12 dB per octave rolloff of the R2C? What word would you then have in reserve for describing the much steeper rolloff of the R3 below its port tuning frequency?

The difference between them in 40Hz is about 8 dB. If the word "plummet" if offensive to you then you can call whatever makes you feel better. The R3 plummets even steeper but it is after 35Hz, which is very low for a bookshelf and even some tower speakers.

 

[MrPeabody]

The difference between them in 40Hz is about 8 dB. If the word "plummet" if offensive to you then you can call whatever makes you feel better. The R3 plummets even steep but it is after 35Hz, which very low for a bookshelf and even some tower speakers.

From the Merriam-Webster:

Plummet : to fall perpendicularly
: to drop sharply and abruptly

The rolloff of the R2C begins at around 100 Hz, but it is not unusually steep by any fair measure, and as such, the word "plummet" does not fit.

 

[aarons915]

That is true if you think it's a good thing that all of the sound is coming from the same place. I think some people actually prefer the bigger presentation of the sound coming from a taller source.

In the latest Triangle Espirit Antal Tower speaker review, @amirm compared the Revel M16 against the Triangle. He admits that the tonality was better with the Revel, but that he still preferred the Triangle because the sound was emitted from a taller vertical space. With the Revel, the sound sounded too much like a point source, which was enough to make him prefer a tonally much worse speaker. The LS50 would do even worse than the M16 did in that comparison.

May be one of those personal preference things.

In real life sounds, are coming from one place whether it be vocals or where sounds originate from various instruments so I don't know how that could be seen as a negative. I'd also like to add that just because the sound may come from a larger line source doesn't mean it will sound bigger, the LS50 generally sound bigger than most speakers I compare blind, it was very surprising the first time I noticed that. This is evidence to me that the vertical dispersion has a lot to do with how big a speaker sounds and not so much the size of the radiating area. I can't say others are wrong if they enjoy more of a "wall of sound" from a tall multi way tower but I also don't see how that would be a more accurate representation of music, especially with vocals.

I did see Amir's comments about the M16 and I agree that a good 2 way with a waveguide is close to a point source. I also remember he made a comment in the Q350 review that they sounded odd because the sound was coming from the center of the box....which makes no sense lol. I think in some cases what we're use to is what we think is correct and Amir's main speakers are the Salon 2 so maybe anything that gets closer to that line source type of sound is what he thinks sounds correct.

 

[Senior NEET Engineer]

LS50 didn't sound like point source to me. There was too much variation in the tonality at different listening positions.

 

[Ata]

I have auditioned the LS50 OG three years ago and did not like it at all, at that time the R300 was the best bookshelf to my ears.

Fast forward to 2021, I auditioned LS50 Meta and R3 and like them almost equally -- "almost" only because I like bass . In the end I went with LS50 Meta because of the slightly better (subjectively) imaging, my personal preference for quieter listening average SPL of around 75 db which stresses neither speaker, the obvious room modes R3 demonstrated (which tripped even Amir in his original review, recently amended), and the AUD1000 price advantage of the LS50 Meta, which practically paid for the SVS SB-1000 Pro that I had to buy with either speaker to satisfy my bass (and cinema) needs.

 

[tifune]

Seeing as how the Genelec The Ones and Paradigm's Founder series also use this D'appolito style coaxial design, I wonder how would a pair of KEF R2Cs work as vertical mains instead of R3s? Also would there be any deleterious side effects (lobing?) if used this way?

R2c
View attachment 122019

I wondered exactly the same and thusly sent one to Erin for review. I was surprised at how different his results are from Kef's published. Was going to paste some screenshots but the whole article is worth review and I can't link it due to some conflagration - he goes over vertical orientation, grille on (ouch)/off, etc.

My take is, due to having young kids in the house and therefore requiring grille on, I wouldn't use these as front L/R. Remains to be seen if entire R-line suffers from grille-induced comb filtering; trying to get a R5 his way to find out.

 

[Ata]

I wondered exactly the same and thusly sent one to Erin for review. I was surprised at how different his results are from Kef's published. Was going to paste some screenshots but the whole article is worth review and I can't link it due to some conflagration - he goes over vertical orientation, grille on (ouch)/off, etc.

My take is, due to having young kids in the house and therefore requiring grille on, I wouldn't use these as front L/R. Remains to be seen if entire R-line suffers from grille-induced comb filtering; trying to get a R5 his way to find out.

Thank you for sending the unit to Erin! This helps me and many others on this forum make more informed choices.

I also noticed how Erin's measurements are different from KEF's, in particular on axis. It would be great if @MZKM and others pick up on that to get preference scores and EQ, if possible.

In any case, the R2C UniQ FR results look different from those of a couple of speakers with identical drivers, the LS50 Meta and R3.

 

[Ata]

Pierre has the preference scores and EQ already worked out: https://github.com/pierreaubert/spinorama/blob/develop/datas/eq/KEF R2c/iir-autoeq.txt

EQ for KEF R2c computed from ErinsAudioCorner data
Preference Score 4.2 with EQ 5.6
Generated from http://github.com/pierreaubert/spinorama/generate_peqs.py v0.8
Dated: 2021-08-07-14:09:57

Preamp: -4.5 dB

Filter  1: ON PK Fc   804 Hz Gain +1.53 dB Q 0.58
Filter  2: ON PK Fc  1920 Hz Gain +3.54 dB Q 8.72
Filter  3: ON PK Fc 12484 Hz Gain -1.81 dB Q 0.29
Filter  4: ON PK Fc  2041 Hz Gain +1.02 dB Q 12.00
Filter  5: ON PK Fc  2650 Hz Gain -1.30 dB Q 6.00
Filter  6: ON PK Fc  3482 Hz Gain +0.64 dB Q 7.70
Filter  7: ON PK Fc   541 Hz Gain +0.87 dB Q 12.00
Filter  8: ON PK Fc   622 Hz Gain -1.12 dB Q 9.11
Filter  9: ON PK Fc   982 Hz Gain +0.81 dB Q 12.00
Filter 10: ON PK Fc  1782 Hz Gain +0.73 dB Q 12.00
Filter 11: ON PK Fc  4521 Hz Gain +0.49 dB Q 11.28
Filter 12: ON PK Fc  1452 Hz Gain -0.86 dB Q 12.00
Filter 13: ON PK Fc  1140 Hz Gain -0.91 dB Q 12.00
Filter 14: ON PK Fc 12411 Hz Gain -0.56 dB Q 9.73
Filter 15: ON PK Fc  2438 Hz Gain -0.41 dB Q 12.00
Filter 16: ON PK Fc  8684 Hz Gain -0.45 dB Q 12.00
Filter 17: ON PK Fc   708 Hz Gain -0.61 dB Q 12.00
Filter 18: ON PK Fc   448 Hz Gain +0.84 dB Q 12.00
Filter 19: ON PK Fc   841 Hz Gain +0.50 dB Q 10.92

Score of 4.2 is really low considering every other KEF speaker out there. One would do better with a Q150, Q350, LS50 (Meta), with or without EQ, and they would need a sub with any of those.

 

[MZKM]

I'll guess I'll post it here:

KEF R2C Preference Rating
SCORE: 4.1
SCORE w/ sub: 6.9

Sensitivity: 84.9dB (300Hz-3kHz ; spec: 87dB @ 1kHz)
Frequency response: +/- 3.3dB 80Hz-20kHz

 

[HooStat]

Wow. I am totally disappointed. But also very glad that I didn't get this speaker (which I was about to do many times). Center channel speakers with a concentric driver shouldn't be that hard. At least for a company like KEF with so much experience and so many resources.

 

[tifune]

TY - Erin was also kind enough to measure vertical positioning and grille on, is data sufficient to generate below for those 2 measurements?

KEF R2C Preference Rating
SCORE: 4.1
SCORE w/ sub: 6.9

 

[BrokenEnglishGuy]

Yes with two woofers you get more like a point source. The ability to localize the bass in the median plane is not the best, but should be good enough to observe the difference. You also get a more narrow vertical beam in the bass frequencies. Side loops should not be a problem with max 400Hz and about 25cm distance of the woofers.

The two woofer version should be the better option, if there are no other differences with the built.

i used many headphones so the '' coherence '' is very important to me since headphones only use 1 driver. I listened speakers like adam a77x and i found annoying the fact that mids sounds in a different places sometimes, I think the problem is solved with the uni-q mids-high.

i have the r7 and i found them as coherent if not more than my R300.. I way prefer the 3-way from kef than the 2-way, that uniq is not made for lows and the crossover is made for 400hz. For me works wonderful, i like speakers because i like the 3-way from KEFs.

 

[MZKM]

TY - Erin was also kind enough to measure vertical positioning and grille on, is data sufficient to generate below for those 2 measurements?

Vertical improves it slightly: 4.2 & 7.0

You can mentally swap the titles of Erin’s globe plots to see the difference. Horizontal dispersion for instance stays close to Omni much higher in frequency but the top end is more narrow.