Beyond the Ariel

[Lynn Olson]

When theory, measurement, and subjective experience fail to correlate is also when it gets most interesting. That means that there is something waiting to be discovered - now, whether you get to be the lucky discoverer is another question. Some problems are hard, and solutions could be decades in the future. Others are easy, and are just waiting to solved today.

The problem is that you really don't know how hard a problem is until you've solved it. It's like being trapped inside a box, and the instructions for opening the box are printed on the outside. Easy when it's solved, not so easy when it isn't. Curiosity, persistence, and a desire to understand are as good a way as opening the box as any.

[JoshK]

The poles often have their over zealous proselytizers. They annoy me a great deal but I care not argue with them on any level as there are better things in life to spend my worries on.

[Hezz]

Lynn,

You have no need to apologize for your approach and process in designing this speaker system. This thread is dedicated to your project and while it may not be the best solution for others, I have no doubt that it will turn out to be one of the best projects ever made and will appeal to many. Not to mention that it has probably been the single best thread ever encountered in DIY audio for an in depth discussion of speaker based audio and acoustic issues.

All to often good ideas of the past were thrown away because of the politics of profit which is the governing aspect of nearly all corporations and businesses and those at the top who want to own a dis-proportionally large piece of the economic pie.

Corporate America no longer has room for an open ended approach to problem solving as it will not fit into the development cycle and consumer/profit model. Getting gain has overtaken the need for knowledge in all large consumer based areas.

[pdan]

Magnetar has been banned. Seems like Corporate America rules even here.

[Lynn Olson]

Originally posted by Hezz

... I have no doubt that it will turn out to be one of the best projects ever made and will appeal to many ...

I appreciate your confidence, but I must honestly admit my success rate has only been about 50%. Most designers quietly bury their failures, although in a corporate environment (I know from direct experience) the marginal products occasionally make it to market. More often the really good ideas don't make it out of the lab, thanks to institutional pressures and corporate infighting.

In a big corporation, there's such a thing as "being too good", a product so good it would threaten the existing cash-cow product line. This is one of the commonest reasons truly excellent products are never made by large corporations - there's just too much invested in the existing product mix, marketing plan, and sales structure.

This whole "Beyond the Ariel" thing is mostly an experiment to see what is possible. Sometimes I ask for what isn't possible; the only way to actually find out is research the history and properties of the underlying devices, and get an understanding of what they can do and what they can't.

Specifications only reveal the most superficial aspects of the device, since they treat it as an idealized "black box" with a few deficiencies here and there. I prefer to understand the device itself, the context in which it was originally designed, what it was intended to do, and what the inherent limitations of the technology are most likely to be.

A good example are the plane-wave measurements of the 288B that Bjorn Kolbrek has on his page. The anomaly between 2.5 and 3.5 kHz is a little mysterious; what is it? It shouldn't really be there, considering how low the frequency is.

Is it the first breakup mode of the aluminum diaphragm? Maybe. It's definitely too low for the tangential surround to be breaking up - that should be 2.5 to 3 times higher in frequency. Is it a reflection from the rear chamber? Maybe, but the wavelengths aren't right. Removing the rear chamber would answer that question.

This is the sort of thing that needs time-domain analysis by MLSSA or ARTA to find what the source really is. This is a completely different approach than simply equalizing the problem and hoping it goes away - for one thing, if it really is a time-domain problem, equalization is the wrong answer, since it will make the time-domain worse. It's a lot better to solve the problem in the appropriate domain - in short, to see where it is coming from and resolve it.

Now, if the problem turns out to be the first mode of the diaphragm, well, short of switching over to the phenolic version of the 288 (the 290, which is also made by GPA), then notch-filter EQ might be necessary, although not really an ideal solution. It's a judgment call between choosing a more lossy diaphragm material versus equalization - both approaches have their own set of drawbacks and tradeoffs. Lossy diaphragms can be prone to side-to-side rocking, many small breakups instead of several large ones, and higher IM distortion at the bottom of the working frequency range; conversely, notch-filter equalization does not actually "cure" the underlying resonance, it merely withdraws drive power at that frequency, and does nothing for narrowband increases in IM distortion centered at the resonance frequency.

So something as simple-looking as the 2.5~3.5 kHz anomaly can have rather complex underlying mechanisms, and the solutions are not always clear-cut, despite what the measurement fundamentalists might say. The extreme example are Lowthers and AERs, which have pretty scary measurements, but sound better than they have any right to. What makes them sound good aren't the terrible measurements - if that were true, they could be replaced with automotive 6X9 speakers with no loss of quality - but the special aspects of the LOwther/AER driver, such as saturated pole-pieces in the gap, and thoughtful choices of VC former and cone materials.

Measurements are useful, but primarily as diagnostic techniques to improve the drivers and the overall system. For an entire speaker system, they can be fairly misleading, unless you have access to the measurements of the individual drivers without crossovers and equalization.

[chrismercurio]

Originally posted by pdan
Magnetar has been banned. Seems like Corporate America rules even here.

I don't know about corporate america ruling here...

The mod's are very tight about things though. My guess is he was warned multiple times and they finallly yanked him. It is too bad. He has strong opinions, was a good resource and builds lots of stuff.

[Lynn Olson]

Yes, too bad about the banning - his last speaker had a lot of good ideas in it.

What's interesting is that people as different as Gedlee, Romy, and Magnetar all have great ideas, although they are coming from completely different places and have radically different goals. Why people can't just cherry-pick the best ideas is something I don't understand - it's certainly a common practice in the industry, going all the way back to Bell Labs and Rice & Kellogg. Some concepts are patentable, so you steer around those, but a lot of what we do is nothing more than re-discoveries of concepts that are many decades old. Much of the underlying technology has never written down and does not appear in the literature, but can be discovered by talking to the "old timers".

What's nice about this forum is that it mirrors the kind of open-ended, collegial discussions that industry designers have among themselves. DIYAudio isn't about winners and losers, it's about sharing ideas.

[agent.5]

Free Magnetar. How do we stage protect in a virtual forum?

[chrismercurio]

I think you have to spam the mods with PM's....but then....that might get you banned too.

[agent.5]

Originally posted by chrismercurio
I think you have to spam the mods with PM's....but then....that might get you banned too.

Not good enough. We need media coverage.