Physical explanation of amp's break in?

Sean...The design/production philosophy formalized by Taguchi is in general use throughout industry, with the exception of some very small outfits where doing it the old way is a tradition, and this may include most high end audio firms. Good performance can be achieved on a unit-by-unit basis doing things the old way. The consequence of not using the Taguchi approach is high cost, and unreliablity, and that fits high end audio pretty well.

The first example usually cited regarding Taguchi method effectiveness is the camera business in Japan. After WW2 Germany was famous as the maker of the most superb lenses and cameras. They had skilled workers who could grind lenses better than anyone else, and they kept their trade secrets to themselves. Japan realized that they could never match the skill of these workers, and so could not produce cameras comparable to the German product. So, using the Taguchi approach, they figured out which pieces of glass in the lens system were hard to make right, and they designed new lens systems that did not rely on exceptional precision of any of the elements. As a result of this effort Japan became dominant in the camera business.

Example 2 is the superiority of Sony TV sets, vs RCA, GE, Sylvania, etc. I don't think the USA manufactures any TV sets anymore.

I could be wrong ,but, I think some Sony TV's are made in the U.S.A..
Some business practices commonly associated with Japan are actually the result of American minds.
I think that Sean is right about break in. Loudspeakers often sound quite awfull out of the box, then settle in to produce the sound we expect for some time. To borrow an expression/analogy from finance, it is much like the bathtub effect. Initially products may often demonstrates a wide vacillitaing quality, then settles in to perform with more consistentcy for much of its life, and then due to age,wear,etc., starts to demonstrate varying quality (though not necessarily identical to the break in period) till it fails.

Eldar, you're right about the role of Taguchi approach in industry in general, but what should be taken into consideration, is that markets for Sony TVs and hi-end audio are quite different, and so different their consumer's demands. Unlike Best Buy clientele, hi-end audio consumer is interested in "in a margin representing the last few percents of what's attainable", as was mentioned in another thread. For any hobby of this kind (audiophilia, wines, cars, etc) all the joy and satisfaction is expected from those few percents, and not about the rest 97% of normal, mass-oriented, healthy and practical minivan-styled stuff. One of the reasons that cause one to purchase a unit for several thousands dollars is a fact (or at least idea from advertising of that product) that it’s manufactured with “hand-made” approach, sometimes manually tuned – shortly, resulting high price is actually a positive factor here - just opposite to Taguchi direction.

Unsound...We all agree that loudspeakers (and phono pickups) improve with use as the elastomers soften up.
As I mentioned, Taguchi was a disciple of W Edwards Deming, whose genius was not recognized here.

Dmitrydr...If there is a "high end" TV, it's Sony. You pay more for a Sony, and it performs better. But, they don't tweek each unit to make this happen. They tweek the design and the productiuon process so that EVERY unit that comes off the line is superior.

Everyone...How is the weather out your way?

Eldartford, I meant a difference between even very good mass-market product and hi-end. Among TVs I would think about Pioneer in this group, but not Sony. It's not just a matter of "better quality-higher price", it's matter of different markets: I guess under normal circumstances you don't come to Best Buy for your next amp upgrade? If you build a graph Price=F(Quality) on some stage it will go up much faster then before, i.e. a little increment in quality will cost much more money then similar increments before. Here comes audiohile's market. If you don't tweek the product, it'll be declined: as per the same graph, for normal practical approach it's worth to go a little bit down in quality for quite significant difference in price.