You may be thinking of XRK971 and his multiple tests of wide band drivers (some of which I skim read).
He did not use EQ, so the quality factor being tested was (essentially) how smooth the driver's raw response was.
The midrange (400Hz to 7kHz) tests was JonBocani:
World's Best Midranges - Shocking Results & Conclusions.
I'm too lazy to skim that thread to see how the tests were done, but, since he was using EQ, I assume the quality factor being tested was (essentially) whether or not the driver had serious breakup modes before 7kHz.
Distortion is horrendous for todays standards.
There is energy storage in passband visible in waterfall.
Frequency response linearity is mediocre.
3/5 is generous - even disregarding the price.
Cheers for the analysis, Zvu! When you say "energy storage in passband" Do you mean the humps at 700hz and 2.5ms?
I run the eton 3-212 mid @ 320hz LR4 to 1700hz LR6 and am just wondering how it stacks up. The mid is -16.5 db down from the 10" woofer to get things to match. I am not seeing any distortion above the noise floor at my listening levels so I guess that is one bullet dodged. I do see a resonance around 700hz like hificompass and wonder if doing anything behind the cone will help. I am going to make a 700hz tuned plate and stick it behind it just to see what I can measure.
I run the eton 3-212 mid @ 320hz LR4 to 1700hz LR6 and am just wondering how it stacks up. The mid is -16.5 db down from the 10" woofer to get things to match. I am not seeing any distortion above the noise floor at my listening levels so I guess that is one bullet dodged. I do see a resonance around 700hz like hificompass and wonder if doing anything behind the cone will help. I am going to make a 700hz tuned plate and stick it behind it just to see what I can measure.
There is that thing going on above 2kHz, and can be seen in waterfall. Instead of having gradual fall without abrupt changes, it has this:
What's visible at waterfall under 1000Hz could very well be a measurement artifact. To be sure what you see is real, much longer gating should be used and/or observed peculiarity on impedance plot. There is some mild hill at that frequency at impedance plot so maybe it is resonance.
Driver of that size used at such a narrow passband has much unexploited potential. For a 3" midrange i'd always opt for >3kHz crossover point.
Fact that your midrange is attenuated 16.5dB doesn't mean much. Your midrange will still have 1% HD3 when playing at 94dB loudness and i'm keeping a blind eye for HD2 (which i shouldn't actually). Not to mention very high HD5 and HD4.
I see no real advantage of Eton 3-212 over Vifa TC9. Vifa is better at everything but the sensitivity - it's 2dB less sensitive that Eton, which is understandable given that it uses Fe magnet compared to Nd used in Eton.
What's visible at waterfall under 1000Hz could very well be a measurement artifact. To be sure what you see is real, much longer gating should be used and/or observed peculiarity on impedance plot. There is some mild hill at that frequency at impedance plot so maybe it is resonance.
Driver of that size used at such a narrow passband has much unexploited potential. For a 3" midrange i'd always opt for >3kHz crossover point.
Fact that your midrange is attenuated 16.5dB doesn't mean much. Your midrange will still have 1% HD3 when playing at 94dB loudness and i'm keeping a blind eye for HD2 (which i shouldn't actually). Not to mention very high HD5 and HD4.
I see no real advantage of Eton 3-212 over Vifa TC9. Vifa is better at everything but the sensitivity - it's 2dB less sensitive that Eton, which is understandable given that it uses Fe magnet compared to Nd used in Eton.
Last edited:
Thanks!
@Zvu, the glassfiber cone definitely looks to be the better material to have close to a tweeter. I can say for sure that my ETON mid does not like my tweeter playing 2-4khz next to hit. The mid rings in my decay. So +1 to the 10f for being the better mid to have close to a tweeter. I am a newb and thought my XO would solve all the problems but did not think of the tweeter playing off the cone.
Just to find something to put in the "Plus" column... I think the 3-212 looks better when it comes to the decay after 1ms and between 300hz and 1000hz. The attached image has the inverted 3-212 on top of the 10f graph. Black lines = 10f , Pink lines = 3-212. The scales do not match. The eton's scale is -2db different at the top and -4db quieter by the time you get to the bottom of the scale. You can see that the black lines stay above the pink after 1ms in my red zone. This graph is telling me that the 10f rings between 300hz-1000hz more than the 3-212. Does 1ms extra decay and +10db of extra ringing mean anything? I dunno. I will probably buy a 10f to compare.
Attachments
What I suspected: decent linear behavior, poor non-linear behavior, and despite all that: sounds really good.
Too often a rather poor objective performer just sounds better than many drivers with significantly better objective performance. You just can't go purely by objective performance.
Too often a rather poor objective performer just sounds better than many drivers with significantly better objective performance. You just can't go purely by objective performance.
The 10F is flat at 30° till 9K/10K Hz and flat to 15K Hz at 15° off axis.
If one accepts the phase and center to center to be less important beyond 5K/6K, then maybe such little good 3" could work with a 6dB or 12 dB XO around 9K Hz to let the 3" working on 4 to 5 octaves ...
Trade offs, but I am going to try a LR12 at 4K hz with a center to center of 1.2 wave length and 6BUT around 8 to 9 K with a bad C2C but with a little Neod. Tweeter having a very small diameter package : around 6 cm diameter or a little AMT to make its lower diag. edge nearer from the 3" center than a round tweeter...
If one accepts the phase and center to center to be less important beyond 5K/6K, then maybe such little good 3" could work with a 6dB or 12 dB XO around 9K Hz to let the 3" working on 4 to 5 octaves ...
Trade offs, but I am going to try a LR12 at 4K hz with a center to center of 1.2 wave length and 6BUT around 8 to 9 K with a bad C2C but with a little Neod. Tweeter having a very small diameter package : around 6 cm diameter or a little AMT to make its lower diag. edge nearer from the 3" center than a round tweeter...
The Eton 3-212 has a silky smooth amplitude response, and has excellent dispersion / low directivity- even at 60 degrees off axis out to 2Khz.
All the harmonics fall into the treble range so will either be difficult to discern or masked by any tweeter that will be playing 40dB louder than it.
To me clear it’s we overrate our human ability to discern harmonic distortion. For years objectivists complain about ribbon tweeters having poor distortion compared to the best domes, but they make up for it in amplitude response and beat them in dispersion. And they can sound wonderful when crossed appropriately.
For years I had a speaker with the Fountek NeoCD3 and crossed over at 3.5Khz I could not detect any “distortion” even though it is was at least 10dB more harmonic distortion than the midrange it was paired to.
I still go for the drivers with the best harmonic distortion performance, but I prioritise linear distortion performance first. Lots of people love the AudioTechnology drive units or Aurum Cantus AC130F1 or GR Research M130- super smooth amplitude response and only average non linear distortion.
All the harmonics fall into the treble range so will either be difficult to discern or masked by any tweeter that will be playing 40dB louder than it.
To me clear it’s we overrate our human ability to discern harmonic distortion. For years objectivists complain about ribbon tweeters having poor distortion compared to the best domes, but they make up for it in amplitude response and beat them in dispersion. And they can sound wonderful when crossed appropriately.
For years I had a speaker with the Fountek NeoCD3 and crossed over at 3.5Khz I could not detect any “distortion” even though it is was at least 10dB more harmonic distortion than the midrange it was paired to.
I still go for the drivers with the best harmonic distortion performance, but I prioritise linear distortion performance first. Lots of people love the AudioTechnology drive units or Aurum Cantus AC130F1 or GR Research M130- super smooth amplitude response and only average non linear distortion.
Last edited: