World's Best Midranges - Shocking Results & Conclusions.

I fully agree on the paper vs metal midrange conclusion.
Looking through your homepage just now. And your findings look very similar to mine.
All drivers breakup at some point - more or less. I cross at 2Khz so that exactly the harmonics that you describe, are kept at bay, cause mine breakup at around 10Khz - leaving both 2 and 3 harmonics beyond the cross-over point. By this - the "problems with hard cone/domes - are actually not a problem anymore 🙂

I did find that my Accuton had a breakup at 4800hz - and with a cross-over at around 2Khz, still left some distortion within the audible area. By notching the break-up before applying the filter, the distortion did lower and improve overall sound quality. Having a driver that is "clean" within an octave - both over and under the required range - does tend to make it sound better and measure better too - also when combined with the rest of the drivers - read: tweeter and woofer.
A DSP is a very perfect tool to hit that problematic spot - very accurately, so that the best possible compromise is reached - cause nothing is perfect - but nothing is better than the tools/methods used either.


It has been spoken so often - that the Revel Salon 2 and JBL M2, sound so alike because of their almost equally even and smooth FR and dispersion. But but..... if you start pushing either of them and understand and use them for their true purpose - especially when it comes to room type and how loud you play. Then I believe we all would quickly find bigger differences.
It kind of reminds me about the rave about perfect phase control in FIR filters vs IIR. It might be heard in controlled environments with special test tones. But under normal circumstances with normal music... not so much.
 
I've tried many small midranges ... They do NOT sound the same.
I don't think that was the point. The point I think was that, in a double blind (ABX) test, midranges could be manipulated so that they all sound quite the same within obvious boundaries. And within sound quality aspects that good reproduction requires. That all in front of an audience consisting of various test persons with reasonable hearing abilities.

Of course such an approach is questionable from a DIY audio enthusiast's point of view. But OTOH we have to accept we do not practice science here, but merely a hobby. Or a trade (also not that scientific).
 
I don't think that was the point. The point I think was that, in a double blind (ABX) test, midranges could be manipulated so that they all sound quite the same within obvious boundaries. And within sound quality aspects that good reproduction requires. That all in front of an audience consisting of various test persons with reasonable hearing abilities.

Of course such an approach is questionable from a DIY audio enthusiast's point of view. But OTOH we have to accept we do not practice science here, but merely a hobby. Or a trade (also not that scientific).
Ok... I can agree with that. But I simply found it difficult to accept that I should be able to live with the slightly more dull sounding soft dome/cone - because one test tried to point out that it all could sound almost equal if hammered in place.
I would argue that there is more to it - not necessarily more complicated - but just a little more nuanced than simple EQ-power.
When it comes to price pr. unit... then I fully agree that we do not necessarily need to pay top dollar to get awesome sound - just look at Kii3 - cheap drivers but nifty design 😀
 
I don't think that was the point. The point I think was that, in a double blind (ABX) test, midranges could be manipulated so that they all sound quite the same within obvious boundaries. And within sound quality aspects that good reproduction requires. That all in front of an audience consisting of various test persons with reasonable hearing abilities.
Of course such an approach is questionable from a DIY audio enthusiast's point of view. But OTOH we have to accept we do not practice science here, but merely a hobby. Or a trade (also not that scientific).
I suggest to trust John Curl ,ABx test only you 😛
 
1.5 db is well shown to not be a sufficient threshold for flatness (as far as a comparison excluding f.r.) you should be within 0.1db ideally.

to show this yourself, setup some ABX sofware and put a 1.5db eq bump somewhere, you'd be able to reliably tell the two apart


Depends on the Q-value. With a Q of 15, you wont hear 1,5dB. But with a Q of 0,5 - you'll easily hear 1,5dB. So I kinda agree with you - but it depends 😉
 
Lol. I wouldn’t dare to claim such a thing. And for the record: I am not disqualifying Curl, I only state his approach isn’t scientific. Funny thing is people often challenge scientific methods with outcomes they don’t like and embrace those that fit their opinion.

As to ABX: it really seems to be the only way to back up or reject a hypothesis about the perceivability of certain signals. I don’t see anybody bring up another scientific approved method. Of course you can reject such a thing. But where does that leave you?
 
Last edited:
You said..

"this is nonsense. He has a vested interest, he can't escape it even with the best of intentions"
You already have, you just infer that they are not clever enough to realise that they are doing it.
Anyway, this is DIY Audio, those experts don't sell kits, or parts. Most of their companies products are outside the majority here's budget anyway. That's all from me.