Interesting midwoofer driver comparison

**5th element** · 10 September 2014, 18:23 Wednesday

That has to be one of the worst conducted tests I have ever seen. Nothing is held constant. Nothing is properly measured, no crossovers are properly optimised or implemented. They talk about things that don't matter at the same time as not understanding them. They also used fixed xover frequencies and slopes for the low passes...it's like some newbie playing with a minidsp and thinking he's arrived at audio nirvana by dialing in a textbook 2nd order LWR at 2.5khz.

I am not surprised that the results varied as much as they did, some designs are going to work better than others when you're leaving it up to chance. We all know how vastly different the upper end of 6.5"drivers can be.

I am, again, not surprised that some manufacturers chose to opt out as the testing is hardly fair, or actually useful in any way.

It's a shame because they could have made something of this if they'd done it properly.

**JonMarsh** · 10 September 2014, 20:16 Wednesday

well, for the way car audio is done these days, the test methodology makes sense- in a somewhat barbaric kind of way- they do just use crossovers with standard slopes and they don't do driver EQ in the crossover- because the electronic crossovers used don't support that.

What is curious or puzzling, is that after equalizing levels, there are the differences quoted for things like mid bass impact and definition that are in a frequency range that I wouldn't expect to be much affected by the things they didn't do right in our mind. Of course, I also have to note that for the drivers I am familiar with that they did rate as Tier 1, (the Dynaudio and the Scanspeak), I think those are first rate drivers anyway, and some of it I would pin on motor linearity and low distortion. Other than that, it's hard to imagine two drivers built more differently than the Illuminator and the Dynaudio (the latter being more like the new Esoteric series, or even the HiVi D6.8.

I've known a few people who check out new drivers just by hooking them up and listening to them full range.... and obviously, drivers with high Q breakups aren't going to be well received under those conditions!

**5th element** · 10 September 2014, 23:03 Wednesday

My problem with the subjective reviews was that they started off pretty quickly by directly contradicted what they'd said during the introduction. Or when I know a driver that measures superbly ends up in the lowest tier, I am then not going to give their subjective impression much weighting when clearly their implementation is going to be what caused the issue.

I didn't read everything in too much detail because I find baseless subjective prose difficult to stomach, so I skimmed a lot of it, but I don't think there's any qualification in there at all with respect to why their testing is flawed.

I mean I've just put together an in car system for my brother using some Dayton Audio drivers and it works very well, but you are severely limited in what you can, and cannot achieve (because of the environment). I mean I measured the ND140 and ND20 on a flat baffle and then designed a crossover to have them hit certain target slopes. It'd be virtually impossible to measure them in the car, due to all the reflections and on listening it was clear that there were issues that I would find difficult to live with (I have hyperacusis ;/), but he thought it sounded a lot better. The truth is it did sound a lot better than the stock system and a bit of EQ did easily sort out the issues that I had (mainly tweeter issues due to massive diffraction from nearby surfaces - go figure!). Luckily he feeds the system music via his smartphone that has some EQ capabilities built in, so all isn't lost!

I get that the car industry does tend to just whack an electronic xover on things, or driver sets come with fixed passive xovers designed by the manufacturers. There's not a lot else you can expect them to do mind you. My problem with this is that small variations in the frequency response have a profound effect on the way we perceive sound.

If you compare Zaph's measurements for the W18, the ND430 and the 8 ohm version of the paper illuminator he measured you easily see where their testing really failed. First of all their idea for ensuring issues on the low end wouldn't occur, by using a fixed size large box, is just a waste of time. What they needed to do was use an LT on every driver set to make sure that they all had identical low end roll offs with identical f3s. The illuminator, just by this standard, has significantly more output in the upper bass and lower midrange due to it's lower resonance and relatively high Qts. The ND430 has a Qts of 0.32 with an fs of 66 Hz whereas the Illuminator has 44Hz and 0.55 by Zaph's measurements. The fact the ND430 sounded a bit thin and lacking comes as absolutely no surprise, but this isn't any fault of the driver, it's because of their implementation. The other aspect is that both the ND430 and the W18 have rising responses from the lower midrange into the top of their passband, the illuminator does not. Yet again they thought that both the ND430 and W18 were thin sounding and lacking in body etc...well what a surprise.

I mean sure, they used some measurements to try and get things as close to 'similar' as they possibly could with their equipment, but we both know that very subtle frequency response differences, say a 2dB general downward sloping trend vs flat vs slightly tilted up can have a profound effect on our enjoyment and subjective impression of a system being full bodied or more thin on the ground.

The reason why I dislike the test is that it gives the W18 and the ND430 a bad name without any explanation as to why. They give the impression that their test and implementation is fair and valid, where in fact it is the root of all their problems. Certainly there is absolutely nothing wrong with either of the W18 or ND430, they are stellar performers in the way they measure, somewhat better than the Illuminator in fact, at least for true operation with a sub. Yet here we are with people saying with some authority that these drivers sound bad, which are realistically, baseless comments due to the way their testing was done.

It's like they made excuses at the start for why they couldn't ensure that more parts of the system were matched, didn't bother to elaborate upon what differences these variations would actually have on the perceived sound quality and then wrote drivers off based upon their flawed testing methodology. If they'd said that these drivers likely sound thin because their T/S parameters give them less output through the lower/upper bass and lower midrange than others, then I'd have been okay with it (because then the reader would know how to compensate for it if they know how to). But all I really see the unjust slandering of excellent drivers due to the inability of the testers to set up a fair comparison.

Besides, don't most serious car audio guys nowadays use a 6" kick panel up to around 300Hz with a very small mid and tweeter up top to handle 300-20k? Like a Scan 10F + neo XT25? Then sub in the boot?

I've known a few people who check out new drivers just by hooking them up and listening to them full range.... and obviously, drivers with high Q breakups aren't going to be well received under those conditions!

Yeah I've read of people doing this too and it makes me cringe. I have no problem with people doing this and commenting that a driver with a well behaved cone sounds pretty decent vs ones that don't. It's when they completely write off the drivers that need special attention because of the bad 'filterless' audition that gets my goat. I'm like...what on earth are you doing you moron, of course it sounded bad anyone could have told you that without even listening, put a filter on it and try again! But no, it's bad, so out it goes.

Interesting midwoofer driver comparison

Interesting midwoofer driver comparison

Comment

Comment

Comment