[pianotech] key leveling with a curve

[Susan Kline]

  On 10/14/2010 6:08 AM, Horace Greeley wrote:
>  P.S. - But seriously, folks, if anyone out there _does_ have some 
> _substantive_ documentation, I really would love to see it...at this 
> stage, one less mystery in life would be most welcome...hg

Absolutely no documentation, whether substantive or smoke-and-mirrors, 
of any kind ... but I can tell you why I sometimes have used a small 
(subtle, you could call it) crown when leveling ordinary new sets of 
keytops on garden-variety pianos. I do it (when I do it) by having one 
side of my straight edge just slightly concave (from warping) and the 
other not. If I use the concave side, I reverse it a few times end for 
end, so that I'm assured of symmetry. And one can see a lot by parking 
one's eye at either end of the keys (in turn) and looking straight 
across the fronts, which I'm doing anyway by the time I get down to the 
yellow-.001"-tissue stage.

I'm sure we're all aware of something called "entasis" ... if the keys 
are (pardon me) dead straight, they look like they are not straight, but 
subtly sunk down in the middle.

Also, if one has replaced the b.r. punchings, (well, with the fluffy 
woven ones we used to use) they will probably pack down over time just a 
fraction more in the middle register than on the ends. Hence, if one 
puts a very small crown in the key leveling, it will probably end up a 
little closer to level a year or two later.

At the concert level, I found Al's Steinway-idea of uniformity of keydip 
to be logical.

Whether even the most picky artists will be bothered would seem to me to 
be dependent on how much crown one decided to use; but also on the 
artists' expectations about the venue, formed before they arrived. One's 
life is much easier if artists expect instruments (tuning, regulation, 
voicing) to be worse than they turn out to be.

Susan Kline

-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://ptg.org/pipermail/pianotech.php/attachments/20101014/231362ca/attachment.htm>