KEF Q100 Speaker Review

[Frank Dernie]

I find that minimally mic'ed recordings sound a lot more realistic than stereo mixes.

In my experience too.
The downside is a bit more noise and, if you think it is a downside, inability to do much dicking about with the mix

 

[JohnBooty]

Finished testing the R3. I may be getting closer to correlating my listening results with measurements. As to scores, I already mentioned some of my uneasiness about them in an earlier post.

I have seen people elsewhere trashing ASR's speaker reviews because of the preference scores.

Folks (including many on ASR itself) fail to understand that those scores are not an "official" part of Amir's work. I'm not really sure how they manage to have this misunderstanding, yet somehow they do.

Myself? I am a big fan of those scores because I understand what they do (and don't) represent. I'm thankful to @MZKM for his work in producing them. However, given the misunderstandings, perhaps @MZKM might consider adding a note to his preference score posts indicating that the scores are not really an "official" ASR thing?

 

[QMuse]

However, given the misunderstandings, perhaps @MZKM might consider adding a note to his preference score posts indicating that the scores are not really an "official" ASR thing?

I don't agree. This score is not a private affair of @MZKM but his immplemetation of Olive's coring method that has been scientifically proven. Untill we have a better scoring system I say we stick with this one and try to improve it. It would help though if Olive would join this forum and enter discussion on that topic as Toole used to do.

 

[hardisj]

I was thinking this a couple days ago when Amir had posted the score himself. I get that it’s science (hey, I’m an engineer who works with missile systems) but I also wonder how the whole inclusion of objective data + subjective feedback + preference score may muddy the waters a bit. At first glance to an outsider it may seem a bit perplexing to have the same source sharing all those different bits of information. (However, logically, I don’t see an issue with it; other than personally I’m not sure I would want to get in to subjective evaluations without a good bit more context and specifics that show I have in some way standardized those evaluations as well)

 

[Francis Vaughan]

I don't understand if you are agreeing or I disagreement with my post.
Would you agree that for the stereophonic trick to work at its best then a point source would be ideal?

Probably neither agreeing or disagreeing. Mostly making a related point.

I don't think a point source has much to offer really. Its has some advantages in controlling the 3d sound field, but these are not special. I would consider a line array as potentially superior. The main trick used in fake spatial effects in a stereo mix is simply pan-potting a single mic'ed source to a point. That is enough to make the ear lock onto a sound. Phase information isn't much used. My beef is that this is just a fake party trick that has little to do with realistic representation of a performance, yet some of the golden eared wax ecstatically about such things as if it is some part of audio nirvana.
Years ago such simple mic setups as the Decca tree were perfected, and little has happened to improve on them.
There are some interesting issues due to the speed of sound that are not mamanged properly in heavily mic'ed setups, and this is, I suspect, another reason that they sound wrong.

 

[Francis Vaughan]

Olive's coring method that has been scientifically proven.

A nitpick. But the scoring method has not been scientifically proven. There is one paper based upon 70 speakers. There has been no peer reproduction of the results. What the method is is a set of parameters derived from experiments performed by one researcher. No validation, hardly a falsifiable theory, not yet science.

IMHO, the scoring system contains some interesting results, but the parameters created for the system are at best very loose. The very wide discrepancies seen for speakers in the middle of the score must give some pause for thought. The number of parameters used to fit the experimental data is also rather high, which leads to more questions.

What Amir is doing will very soon surpass the data used in Olive's work. No blind preference data, but soon more speakers analysed. There will be more insights. Right now the scoring method is an interesting side issue. But no more.

 

[tuga]

I think Linkwitz said that dipole speaker's imaging is "realistic", whether it is perfect or not is not known. The sound comes from the room and walls not from two point sources like box speakers, that is what he means "realistic". I never heard of dipoles, so does not have any comment on perfect imaging.

Stereo uses two speakers to trick the brain into believing that "images" instruments are being created between those speakers (real stereo anyway).
I don't understand what he meant by "realistic" but can't see how using wall reflections can produce a sharper image than direct sound.

 

[tuga]

Probably neither agreeing or disagreeing. Mostly making a related point.

I don't think a point source has much to offer really. Its has some advantages in controlling the 3d sound field, but these are not special. I would consider a line array as potentially superior. The main trick used in fake spatial effects in a stereo mix is simply pan-potting a single mic'ed source to a point. That is enough to make the ear lock onto a sound. Phase information isn't much used. My beef is that this is just a fake party trick that has little to do with realistic representation of a performance, yet some of the golden eared wax ecstatically about such things as if it is some part of audio nirvana.
Years ago such simple mic setups as the Decca tree were perfected, and little has happened to improve on them.
There are some interesting issues due to the speed of sound that are not mamanged properly in heavily mic'ed setups, and this is, I suspect, another reason that they sound wrong.

Wouldn't a line array create 6ft high images of say a violin, such as what one gets from a large Magneplanar?

 

[tuga]

Linkwitz said that dipole speaker's sound comes from the walls, not from two point sources from stereo speakers. Dipoles also have better off axis if setup is right, so the sweat spot is bigger. I think what Linkwitz thought on realistic was that dipole speaker feels like the musician is playing in the room and the sound bounces and fills in the room. Everyone will have different opinion on what is realistic.

Room reflections will make the sound more enveloping but they will also reduce image sharpness.
As for musicians playing in the room, that has got more to do with the recording than with the playback.

 

[Francis Vaughan]

Wouldn't a line array create 6ft high images of say a violin, such as what one gets from a large Magneplanar?

I've never noticed tall images from maggies. I had a pair of Tympani IV at home for a few months many years ago. The physics of extended sources is a bit counter intuitive. Maggies are very interesting speakers in many ways. If you have a tall sonic image I would be looking at your floor and roof acoustics. No speaker lives in isolation. (The room I had them in had a tall raked ceiling.)

I think the point I am trying to make is that almost all of these pinpoint stereo images are not realistic. Some people love them, some people are ambivalent. That is personal taste. But what one cannot do is claim they are representative of some acoustic truth or realism. Linkwitz clearly regarded them with ambivalence, and sought an aural experience that was a better match for his experience of reality in musical performance. YMMV.

 

[thewas]

Large loudspeakers usually have higher directivity which leads to rather smaller but more sharper imaging than smaller ones.*

*Of course only ones with smooth directivity and no diffraction problems which is counteractive to good imaging.

 

[ROOSKIE]

We are agreeing to the extent that no one is saying Amir's amp is distorting. That was the only agreement I was implying.

As for the rest, I get what you're saying, but qualitative data is still data (albeit with a big asterisk) . Get enough of it saying the same thing and that warrants further quantitative analysis to figure out what is going on. Saying "everyone is imagining this distortion" seems a bit hand-wavey to me.

His amp may be distorting. There is no way that there is no way, it is possible.
If he is listening loudly in a large room with only one speaker driven to high volumes I would be surprised if 100watts is always enough.
As well that little driver would be working very hard alone, at least run two so the high volume levels have a helper.
I think what people are really discussing is that the subjective portion of the testing here is of poor quality and in direct conflict with the high quality off the measurements. It is degrading the overall presentation of an industry standard test site.
His subjective reviews are some of the least thorough - essentially amazon ratings and the measurements on the other hand are outstanding.

So much good work, yet there must be a quality subjective component. I don't know the solution though.
I do love this site.

 

[Ron Texas]

How do you measure magic? Coherence between tweeter and woofer.

The problem is that the frequency response on the axis is just one more factor. For the vast majority it seems that it is the only one that matter. Same as SINAD with amplifiers, wrong. Unfortunately the reality is much more complex and it is not so easy.

Perhaps without the magic part, this gets discussed by others around here. Problems like dispersion, diffraction, cabinet resonances and lack of power handling can't be fixed. Flat response on axis can be EQ'ed. The research indicates a preference for for flat response on axis and even dispersion. There is a correlation but there are panel speakers which measure poorly and people love them. SINAD thing is overdone because there is a point beyond which it doesn't make any difference.

 

[QMuse]

A nitpick. But the scoring method has not been scientifically proven. There is one paper based upon 70 speakers. There has been no peer reproduction of the results. What the method is is a set of parameters derived from experiments performed by one researcher. No validation, hardly a falsifiable theory, not yet science.

IMHO, the scoring system contains some interesting results, but the parameters created for the system are at best very loose. The very wide discrepancies seen for speakers in the middle of the score must give some pause for thought. The number of parameters used to fit the experimental data is also rather high, which leads to more questions.

What Amir is doing will very soon surpass the data used in Olive's work. No blind preference data, but soon more speakers analysed. There will be more insights. Right now the scoring method is an interesting side issue. But no more.

Well, in that case I stand corrected.

 

[QMuse]

Large loudspeakers usually have higher directivity

This seems like quite a long shot to me. What would be the logic behind this claim? Pretty much every good small speaker can be converted to large floorstander - make box larger and stiffer, replace woofer with larger unit, add another woofer if needed and there you are. Tweeter and MF driver remaine the same (if there was MF driver), and there you are. So where in this process you think directivity gets narrower/larger?

 

[thewas]

What Amir is doing will very soon surpass the data used in Olive's work. No blind preference data, but soon more speakers analysed. There will be more insights. Right now the scoring method is an interesting side issue. But no more.

Imho they can't be compared because they are pure measurements without any correlation to any controlled experiment subjective preference rating. That doesn't mean that I don't value the measurements here, I actually love them because with the knowledge we have got from Toole we can nicely see what is good and what isn't and that even without any calculated scoring.

 

[thewas]

This seems like quite a long shot to me. What would be the logic behind this claim? Pretty much every good small speaker can be converted to large floorstander - make box larger and stiffer, replace woofer with larger unit, add another woofer if needed and there you are. Tweeter and MF driver remaine the same (if there was MF driver), and there you are. So where in this process you think directivity gets narrower/larger?

That usually larger loudspeakers have larger drivers (exceptions prove the rule ) and also due to their bigger baffle their baffle step frequency at which they become directive is lower.

 

[maty]

Perhaps without the magic part, this gets discussed by others around here. Problems like dispersion, diffraction, cabinet resonances and lack of power handling can't be fixed. Flat response on axis can be EQ'ed. The research indicates a preference for for flat response on axis and even dispersion. There is a correlation but there are panel speakers which measure poorly and people love them. SINAD thing is overdone because there is a point beyond which it doesn't make any difference.

I disagree. The frequency response must have few oscillations, if possible +/- 2 dB and also the spinorama must be excellent. If the second is, it will be more feasible to equalize to solve the small defects that we detect. In other words, spinorama is much more important because the above is assumed in a quality speaker.

Explain to me how you measure coherence on a coaxial that detects clear differences between the 5.25" and the 6.5" from KEF Uni-Q.

 

[maty]

In the impedance and phase curve I only see a slight resonance at 750 Hz if I enlarge the image. So the big improvement I appreciated when adding mass + damping should not happen.

 

[QMuse]

That usually larger loudspeakers have larger drivers (exceptions prove the rule ) and also due to their bigger baffle their baffle step frequency at which they become directive is lower.

You are a young man, you should not live in the past!

Check how it is done these days with modern designs, same width, same drivers:

Kef Reference 1 Bookshelf:

Kef Reference 5 Floorstander: