Personnaly, subjectively, i ''FEEL'' a difference between many of them, even EQd and Level-matched.
Problem is: i cannot prove that i'm able to identify them! 😱
... but anyway, even not-blinded, once EQd the difference i ''feel'' is not much. For sure it's a lot less than what i expected.
Maybe you should focus with very clean and good reccordings with a big amount of harmonics then :
- try to choose the ones which are the more informativ about détails on the time scale ?!
- then try to check if they are the most appreciate by the listeners
- then measure it to see how the 5 harmonics curve and the shape of the time decay/spl Spectrum in ms/db
- then try to see if there is a correlation ?!
Myself I find than often high mid and treble with speakers are too much informative in relation to the lower Fhz range ! It's not about the sound stage and the reverbs of the room only imo. Btw it's very not a problem with Rock-Pop reccordings but more with voices and acoustics instruments.
One important factor as well is the quality of the source (the player) : only the best gives you this informativ low mid and bass ! Sometimes (well : often imho) it's not the speaker or the driver but certainly the quality of the source if the designer made the XO filter correct !
Well here we assume it is as this is standalone driver benchmark with just Gain adjust and I hope in the same Frequency Windows to have a concistency (so if no electronic very stepped filter for the listening test : at least approximative close CE curve and same Sd between drivers !
After it's not forbidden to make subjective benchmark as it's ABX : different Sd, technology of drivers....
I would like to know if the first five harmonics have to be in the good order for better results : the fifth Under the forth, etc ! Or if it is just not important if it is below - 40 dB... WHat is the culpritt : amp, source ???
Last edited:
I enjoy listening to music in a ABX test. It is a MUSIC, damn it! And I am using the same subset of my brain - the one which is between my ears.

just found a pair of Tang bang W3 in my closet 🙂
would be nice to try it as well...
would be nice to try it as well...
An externally hosted image should be here but it was not working when we last tested it.
Few things i'd like to try next (in order of priority/curiosity):
- Voxativ AC-1.6
- TWO identical drivers V.S. ONE of the same model
- a 10'' or even a 12''
- Tang Bang W3
- Visaton B200
we need some fresh out-of-the-box ideas
- Voxativ AC-1.6
- TWO identical drivers V.S. ONE of the same model
- a 10'' or even a 12''
- Tang Bang W3
- Visaton B200
we need some fresh out-of-the-box ideas

One other possible thing to try would be to put a (fixed) tweeter under the Cube, so to reach a 360hz-20khz.
The same tweeter/EQ for all the midranges of course, level-matched as always. Tweeter is fixed, but the Cube rotate to as usual.
Makes any sense ?
The same tweeter/EQ for all the midranges of course, level-matched as always. Tweeter is fixed, but the Cube rotate to as usual.
Makes any sense ?
Ok, time to get serious about the equipement.
Byebye iPad and miniDSP plateamp, let's bring the big boys stuff:
- nanoDigi all-digital DSP
- macMini source
- (2) 1000ASP amplifiers
- Forssell MADA-2 DAC
The real important thing here is to avoid a double conversion AND, of course, to get the best possible digital to analog converter. To my knowledge, the Forssell is the best available on planet earth, followed by the Weiss DAC1.
Now we'll know if the source/DAC is the bottleneck in this test.
Byebye iPad and miniDSP plateamp, let's bring the big boys stuff:
- nanoDigi all-digital DSP
- macMini source
- (2) 1000ASP amplifiers
- Forssell MADA-2 DAC
The real important thing here is to avoid a double conversion AND, of course, to get the best possible digital to analog converter. To my knowledge, the Forssell is the best available on planet earth, followed by the Weiss DAC1.
Now we'll know if the source/DAC is the bottleneck in this test.
ICEpower 1000ASP will be the bottleneck. Those are good modules in general, but not well suited for low-level midrange/treble playback as they lack detail/resolution.
Your milage may vary....
Your milage may vary....
Ok, time to get serious about the equipement.
Byebye iPad and miniDSP plateamp, let's bring the big boys stuff:
- nanoDigi all-digital DSP
- macMini source
- (2) 1000ASP amplifiers
- Forssell MADA-2 DAC
The real important thing here is to avoid a double conversion AND, of course, to get the best possible digital to analog converter. To my knowledge, the Forssell is the best available on planet earth, followed by the Weiss DAC1.
Now we'll know if the source/DAC is the bottleneck in this test.
I think it won't matter much, as long as you use pink noise to dial in the EQ at the listening position. Believe it or not, you're EQ-ing the speaker and room if you do.
If you'd use some type of gating, preferably frequency dependent and EQ based on that, you'd probably be more able to distinguish between drivers, as the indirect sound of each driver (off axis response) would still matter.
Right now you swamp the listening spot with sound, and average that out with EQ. In my opinion that's not a valid way to do the EQ. You could convince me otherwise by posting an IR and Filtered IR, measured at the listening spot, showing the first ~25 ms. If it's clean without showing reflections you prove me wrong.
It wouldn't be that hard to try something different, right?
He's using ICEpower ASX series, which are better than the ASP series. But sure, there are even better amps.ICEpower 1000ASP will be the bottleneck. Those are good modules in general, but not well suited for low-level midrange/treble playback as they lack detail/resolution.
Your milage may vary....
12 people passed the test so far.
Cannot draw final conclusion yet, but it seems people are not able to identify drivers, most of the time.
That alone is a big, big, big surprise to me.
When properly eq'd for flat and with similar distortion profiles, drivers tend to sound the same. Unless there are severe break up/energy storage issues that is (which distortions would show up on the FR and would be impossible to eq out as they are non-minimum phase).
He's using ICEpower ASX series, which are better than the ASP series. But sure, there are even better amps.
could use any icepower models if that's better.
i really doubt the amplifier will make any difference
When properly eq'd for flat and with similar distortion profiles, drivers tend to sound the same. Unless there are severe break up/energy storage issues that is (which distortions would show up on the FR and would be impossible to eq out as they are non-minimum phase).
still, i EQd flat the Dayton RS225 😱
Regarding you selling the Radian PB 950 BeTruextent drivers. You can't eq in a horn that probably should not be used below about 1000Hz. Even if you get a flat frequency response from 360Hz and up. You are selling the winner of your test, without even listening to them properly. Get a larger horn and see how you like them. I would suggest using them from 600-2500Hz and get a tweeter from 2500Hz and up. The horn would need to be 60-80cm in diameter at the mouth. Use horn resp or Autotech horns for inspiration.
Regarding you selling the Radian PB 950 BeTruextent drivers. You can't eq in a horn that probably should not be used below about 1000Hz. Even if you get a flat frequency response from 360Hz and up. You are selling the winner of your test, without even listening to them properly. Get a larger horn and see how you like them. I would suggest using them from 600-2500Hz and get a tweeter from 2500Hz and up. The horn would need to be 60-80cm in diameter at the mouth. Use horn resp or Autotech horns for inspiration.
Rewind,
1. I think you are missing the last dozens of pages of this thread: there is no winner, whatsoever, as of yet.
2. A larger horn couldnt fit in the Cube.
3. A flat EQ is a flat EQ. I get a 360hz-7.2khz all the way within 0.5-1db. It is flat and it sounds flat.
4. On a subjective appreciation, i don't think a Compression Driver (even the famous TADs) is to the liking of everyone. Exotic horn or not. That sure wouldnt be my first choice, i can tell you that..
I would suggest using them from 600-2500Hz and get a tweeter from 2500Hz and up
Just so i am clear to everyone who reads this thread but not from the beginning:
I am using for this test bandpass crossover 360hz to 7,200hz (more or less 4½ octaves) in order to keep a certain level of Listenability.
I KNOW this is not the best band for most drivers, but it's not important in this test, as we just try to identify them from each others.
Question for all:
I'd like to test one of the following:
1. Same drivers but anywhere between 1db and 2db difference
2. Same drivers but bandpass xover anywhere between 1/3 to 1 octave difference
any preferences/curiosity about ?
*same driver = both the exact same model installed in the Cube.
I have few PAIRS available: Airborne, FR10, Alpair, Visaton B200...
I'd like to test one of the following:
1. Same drivers but anywhere between 1db and 2db difference
2. Same drivers but bandpass xover anywhere between 1/3 to 1 octave difference
any preferences/curiosity about ?
*same driver = both the exact same model installed in the Cube.
I have few PAIRS available: Airborne, FR10, Alpair, Visaton B200...
Last edited:
I think it won't matter much, as long as you use pink noise to dial in the EQ at the listening position. Believe it or not, you're EQ-ing the speaker and room if you do.
If you'd use some type of gating, preferably frequency dependent and EQ based on that, you'd probably be more able to distinguish between drivers, as the indirect sound of each driver (off axis response) would still matter.
Right now you swamp the listening spot with sound, and average that out with EQ. In my opinion that's not a valid way to do the EQ. You could convince me otherwise by posting an IR and Filtered IR, measured at the listening spot, showing the first ~25 ms. If it's clean without showing reflections you prove me wrong.
It wouldn't be that hard to try something different, right?
I agree its an obvious problem
- Home
- Loudspeakers
- Multi-Way
- World's best midrange Blind Testing - Need your help.