KEF R3 Speaker Review

[maty]

go Here for this speaker (r3). You can click on a plot and then you can zoom, get tooltips, select 1 graph, print, ...

But... is R3 and not Reference 3, much more expensive! Only $ 11,000.
https://international.kef.com/products/reference-3-floorstanding-speaker

 

[pierre]

But... is R3 and not Reference 3, much more expensive! Only $ 11,000.
https://international.kef.com/products/reference-3-floorstanding-speaker

Thanks Maty, i will correct it.

 

[Thomas savage]

A reminder, personal jibes and insults are not desirable. @QMuse your going to have to change tact in this regard going forward. I know some folks can be irritating but please find another avenue for your frustrations .

cheers

 

[QMuse]

A reminder, personal jibes and insults are not desirable. @QMuse your going to have to change tact ok in this regard going forward. I know some folks can be irritating but please find another avenue for your frustrations .

cheers

If that is related to my RTFM post, I didn't mean it as any kind of insult but as a plain joke.

 

[georgeT]

Sure, if the internal resistance of the new coil is different. An electrolytic capacitor is exchanged for a film capacitor with a significantly lower ESR.
And very important if the values of the components vary within the given tolerances.

If a "normal" film capacitor is replaced by an abstruse expensive film capacitor with exactly the same capacitance, there will be practically no difference.

Was looking this up the other day, stumbled across a hilarious parts express thread where they unanimously agreed that red caps are the only thing that brings a noticeable improvement.

 

[tecnogadget]

Was looking this up the other day, stumbled across a hilarious parts express thread where they unanimously agreed that red caps are the only thing that brings a noticeable improvement.

My god LOL !!! Thats pretty much the antitesis of what this forum tries to achieve.

Tell them to take the “red cap” and enter The Matrix ...Or take the “blue cap” and become ASR members hahaha !

 

[SmackDaddies]

I am really starting to think the 100 to 200 Hz region plays a much stronger role than we think in subjective sound a speaker produces

This.

 

[cleareyes]

How would these compare to neumann kh 120?
I just decided to order a pair, but the r3 seems to be similar in price, without an amp that is.

 

[BYRTT]

go Here for this speaker (r3). You can click on a plot and then you can zoom, get tooltips, select 1 graph, print, ...

Fantastic eye candy visualizations there great thanks the link and hard work, allow me post a copy of your nice radars into a animated comparison R3/M16/8341A

 

[raindance]

I'd suggest, KEF being British and all, that they're tuned for typical British rooms where they will likely be close to the front wall. Indeed KEF bookshelf speakers I have tried seem rather lean unless they are reasonably close to the wall. The other observation I'd make, is that once the bass is brought into balance in this way, they tend to sound rather polite at the high end.

 

[pozz]

The ATC domes are some of the best drivers I’ve personally ever used or heard.

Do you have driver measurements available for them?

 

[ctrl]

Was looking this up the other day, stumbled across a hilarious parts express thread where they unanimously agreed that red caps are the only thing that brings a noticeable improvement.

Inspired by Ethan Winer's "Null-Tester" on the topic "cable sound", I recently tried to transfer his approach to the topic "capacitor sound" by means of analog-acoustic measurements (unfortunately this is a German forum where you have to be logged in to see the pictures).

I also compared a "high-end capacitor" with ancient foil capacitors to see how big the measurable difference is.

If someone is interested in the results, we can discuss this in a separate thread.

 

[tuga]

I am really starting to think the 100 to 200 Hz region plays a much stronger role than we think in subjective sound a speaker produces.

This.

Here's a slice from a Richard Strauss tone poem:

 

[digicidal]

Here's a slice from a Richard Strauss tone poem:

Regardless of how my ears might feel about it... my eyes find it delicious. (And my brain says "needs more subs").

 

[georgeT]

My god LOL !!! Thats pretty much the antitesis of what this forum tries to achieve.

Tell them to take the “red cap” and enter The Matrix ...Or take the “blue cap” and become ASR members hahaha !

I'm sorry, I wasn't being clear, the hilarious part was that after 2-3 pages they concluded that there shouldn't be any difference betwen inductors/resistors/caps as long as they have the declared value. then they started joking for another 5 pages about an expensive red cap sold on parts express as being a clear exception.

 

[MZKM]

It seems that the KEF R3 in-room is NOT following the harman target curve of sloping down. If anything, it peaks back up.

Interesting, could this be caused by more HF reflections due to being in a garage?

 

[QMuse]

Interesting, could this be caused by more HF reflections due to being in a garage?

Posting impulse response would come handy in checking that theory.

Btw, such in-room measurement should always be gated based on the timing of first reflection observed from impulse response. As that has not been done I think we shouldn't be jumping to conclusion based on that measurement. AFAIK we don't even know from which distance it has been taken.

 

[aarons915]

One has to ask a question why that finding is not reflected in a scoring system.

It is reflected in the PIR, you'll notice it's almost the identical curve as the Early reflections in most cases.

 

[napilopez]

Wow that 2m measurement shows pretty good agreement with the prediction; apart from your room showing it's ass in the bass of course!

Yeah, my experience is that by 2M the PIR looks pretty close to an ungated measurement, as expected. The 3M measurement would probably look closer with just one speaker firing and proper toe in.

Thanks for the two comments. I think there are two important statements there.

Firstly, that the listening impression of loudspeakers is not as individual as too many people still believe, although this is repeatedly mentioned in many forums. This is proven by the impressive correlation of the hearing tests by Olive et al.

The second statement is mentioned less frequently, namely that the hearing tests were conducted under very specific conditions and that the calculated formula for determining the hearing evaluation was tailored to precisely these conditions.

In Post#182 I (and many others before me) had listed the test conditions: Mono, large distance to the side walls.
Anyone who has ever tuned the crossover for a free-standing loudspeaker in mono knows that it will sound much too bright in a "normal" listening room in stereo. A speaker that sounds rather boring in mono will most likely sound much better in stereo.




This is an impressive comparison. It seems that the calculated space curve fits better for the artificial test conditions (mono, large distance to side walls) than the "normal" listening conditions in the living room (although a single comparison certainly cannot be generalized).

As mentioned above, there are indications that the artificial test conditions cannot be completely transferred to stereo listening in a living room.

Perhaps the matter has to be approached quite differently. In "Quantifying Sound Quality in Loudspeaker Reproduction" a very interesting approach is used:

Good points all around, I would just caution against thinking that the 3M curve wouldn't match the PIR at this point, as the speakers were not toed in for those compared to the 2M and 1M, which were toed in. I do also have a few old un-averaged measurements of toe in-vs no toe in on the R3 and the toed in ones are a bit brighter in the upper mids/presence but I think the curves might be too "messy" to be useful, even with smoothing.

I've made a note to investigate this further next speaker I measure, or whenever I have the time. Will do single speaker measurements from both my measurement location and the speaker at 0.5M, 1M, 2M, 3M, and 4M and see how the in in-room curve changes.

Will also do some more stereo measurements with toe in vs no toe and see how much that affects the in-room response. My recollection is that the difference is indeed visible, even when averaging over a small listening area, but we'll see when I have the data.

Interesting, could this be caused by more HF reflections due to being in a garage?

Posting impulse response would come handy in checking that theory.

Btw, such in-room measurement should always be gated based on the timing of first reflection observed from impulse response. As that has not been done I think we shouldn't be jumping to conclusion based on that measurement. AFAIK we don't even know from which distance it has been taken.

Unless I am misunderstanding you @QMuse, you do not want to gate the impulse response in this case. Gating it turns just turns it into a quasi-anechoic measurement. Amir was looking to see how the PIR matches the actual in-room curve.

That said, yes, we don't know the measurements distance yet. As shown in my earlier REW capture, the in-room curve will change pretty dramatically with distance. Aside from the image I've posted, I've seen this happen with many different speakers as I always measure them at 2M, 1M, and sometimes closer(such as to see how driver integration changes with distance, out of curiosity). I also measure from my LP. In my particular setup, I've noted:

At just 1M or less, in a relatively unobstructed area like my measurement location, the top third or half of the in-room, ungated curve looks almost identical to the gated measurement. By 2M, the curve has tilted more, with a bit more reinforcement in the bottom half, and more messiness. At further distances, it increasingly approximates the sound power.

 

[QMuse]

Unless I am misunderstanding you @QMuse, you do not want to gate the impulse response in this case. Gating it turns just turns it into a quasi-anechoic measurement. Amir was looking to see how the PIR matches the actual in-room curve.

Aha. In that case he should measure it with RTA pink noise from at least 2m distance using MMM.

EDIT: This is how it looks with my speakers, RTA pink noise from LP (4m) vs gated sweep from 1m distance. Progressive loss of energy vs frequency is a natural fenomena, it happens with all speakers. If you don't see it in your cuvre you haven't move far enough.