FAC measurements on similar pianos

[Robert Scott]

Newton Hunt wrote:

>With a recent SAT, with the FAC feature, he does not _need_ other
>tuning charts, he can make his own, which is better because no two pianos, even
>off the production line consequentively (sp, well, you know what I mean) will
>have different FAC measurements.
>

    Here is something that has always puzzled me.  It seems that anyone
who uses the FAC method is already banking on the hope that there is some
predictability about inharmonicity.  After all, the FAC method only measures
three strings and uses information about those three strings to predict
inharmonicity in the rest of the piano.  If this method works, it is highly
unlikely
that it can be used to beneficially customize tunings on a series of nominally
identical pianos.  The reason I say this is that if there are differences in
inharmonicity between nominally identical pianos, then those differences are
most likely random string-by-string differences that would not be improved upon
by the FAC method anyway.  To believe otherwise is to believe that there are
factors that affect inharmonicity uniformly that are varying from piano to
piano in
a series of nominally identical pianos.  Now I suppose that from one piano
to the
next it is possible that a global change could occur.  They could decide to
switch wire suppliers and string the next piano entirely with wire from a
different
supplier.  This might affect inharmonicity uniformly. But is this the kind
of variation that tuners refer to when they say that each piano is unique?
Or are they really refering to string-by-string variations that do not
correlate across the piano?  Are we talking about unique pianos or unique
strings?

   -Bob Scott