Edgar

But does it sound better? 🙂

Using the "breaking-in a pair of shoes" analogy, imagine walking for 600 hours. At 2 mph that would be 1200 miles, by which time most shoes would not only be broken-in, but worn-out.

It would not surprise me if sample-to-sample production tolerance differences between drivers had more effect than break-in. For the record, that's just conjecture on my part.

I get the impression that modern compression drivers are designed (at least in part) for much greater power handling and excursion than vintage compression drivers (such as my Altec 288). Do you find that to be the case? If so, what compromises were made in the process?

That is torture for a compression driver! Do you also get a lot of infant mortality with this test?

But does it take 200 hours at close to full power to happen?

You beat me to it. We've gone from changes that are "significant" (the best word that I can think of to summarize the various claims about the sonic changes) to "a few percent". So which is it? In my experience, woofers do break-in, and the effect is measurable. Whether it's audible is in the ears of the beholder.

I can't argue with that.

You'll find that many of those early stereo recordings don't mix down to mono well at all, because some of the stereo content was out-of-phase.

A top fuel drag motor only makes about 1000 revolutions between teardowns.

I tried a little experiment with WinISD. First I modeled a K33E in a vented box. It is shown in green on the attached graphic. Then I reduced its compliance by half (equivalent to doubling the suspension stiffness), let WinISD recompute the T/S parameters, and modeled that in the same sized box tuned to the same frequency. It is shown in red on the attached graphic. A doubling of stiffness caused about 1 dB difference in the frequency response.

Very little effect below about 300 Hz. Essentially no effect below about 125 Hz. As expected, given the thickness.

I have noticed that highly efficient (or highly sensitive, take your choice) speakers also handle dynamics better than inefficient speakers. At least, that's how my ears hear it.

Years ago we were using a FFT spectrum analyzer to check the noise floor on a piece of equipment that my group had designed. The SA showed superb noise performance, yet when we listened we could clearly hear an unacceptably high noise floor. It took a bit of investigation, but it turned out that what we heard was impulse noise (a series of discrete clicks occurring often enough that they sounded like noise) in which, by sheer coincidence, the period of the impulses almost exactly matched the spacing between the FFTs. More often than not, the impulses occurred during the intervals in-between the FFTs, so they were rarely seen by the SA. Just one case in which what we measured did not agree with what we heard.

https://www.don-audio.com/Edcor-transformers

I lived in Minneapolis for six years. Never heated my block. Never failed to start.

Edcors are not Jensens, but offer most of the performance for a fraction of the price.

If you've ever owned a pair of real leather shoes or boots, you know that they eventually break-in ... and that they also eventually break-down. We try to minimize both problems in loudspeakers and electronics.

But it starts so much better with the cryogenically-treated wire! 🙂

Try "Commercial & Industrial Power Strips" at Grainger.

As with all engineering problems, the "optimal" solution depends entirely on what you are trying to optimize. If you are trying to maximize headroom and minimize excursion, then your solution may be optimal. If you are trying to smooth low bass response, then Todd Welti at Harman may have the last word: https://www.harman.com/documents/multsubs_0.pdf.

I can see where woofers, which move a considerable distance, would need some break-in, but it happens quickly and the effect is not nearly as large as one might expect. I measured a pair of EVM-10m woofers right out of the box, and then again after a full week of use. The resonant frequency changed only a couple of percent. I have the numbers recorded somewhere, now if I could just find them ...

https://community.klipsch.com/index.php?/topic/191547-odd-noise-in-system/&do=findComment&comment=2494961

Not to mention the cost of liability judgments, should side-effects be discovered twenty years down the road.

Let's discuss the realities of product pricing for a moment. (This has been discussed before, but it bears repeating.) It is a simple fact of life that the retail price for goods like loudspeakers and consumer electronics generally has to be between 5x and 10x the BOM (bill of materials) cost for the product to be commercially viable. That means a $6600 pair of Cornwall IV contains between $660 and $1320 worth of components and materials. The total cost of six drivers, four horns, two balancing networks (each with six inductors, six capacitors, and various other pieces and parts), MDF and/or wood for two cabinets, veneer, glue, fasteners, etc., has to total somewhere between $660 and $1320. That's not a lot to work with for a premium product. Could you build them yourself for less than $6600? Of course. Do you have the knowhow? And the skill?