<!DOCTYPE HTML PUBLIC "-//W3C//DTD HTML 4.0 Transitional//EN">
<HTML><HEAD>
<META http-equiv=Content-Type content="text/html; =
charset=iso-8859-1">
<META content="MSHTML 6.00.2800.1458" name=GENERATOR>
<STYLE></STYLE>
</HEAD>
<BODY>
<DIV><FONT face=Arial size=2>Hi Ric, everyone,</FONT></DIV>
<DIV><FONT face=Arial size=2></FONT> </DIV>
<DIV><FONT face=Arial size=2>Thanks for all the feedback. I =
look forward
to hearing even more!</FONT></DIV>
<DIV><FONT face=Arial size=2></FONT> </DIV>
<DIV><FONT face=Arial size=2>You wrote:</FONT></DIV>
<DIV><FONT face=Arial size=2></FONT> </DIV>
<DIV><FONT face=Arial size=2>> Stanwoods particular curve =
shape is
<BR>> somewhat a subjective choice me thinks... tho he may say
different. He <BR>> can and from time to time does provide =
custom
curves for given <BR>> situations that are not quite this same shape. =
Steinways own default <BR>> curve is somewhat flatter =
btw.<BR></FONT></DIV>
<DIV><FONT face=Arial size=2>Ah hah! Yes, the hammers I have =
are
duplicates for a Steinway D, so that would make a bit of sense. =
The curve
isn't just "flatter," but really quite flat. It's flat as though =
by
intent. I find that interesting. A picture is worth a =
thousand
words. (This GIF, at 9 kB, is pretty economical.)</FONT></DIV>
<DIV><FONT face=Arial size=2></FONT> </DIV>
<DIV><FONT face=Arial size=2></FONT><IMG alt="" hspace=0
src="cid:002b01c48862$f3cfc930$0202a8c0@SarahDell4600" =
align=baseline
border=0></DIV>
<DIV><FONT face=Arial size=2></FONT> </DIV>
<DIV><FONT face=Arial size=2>The jump between 21 and 20 is of course =
the
bass/tenor break, with the bass hammers having more molding to =
them. The
"SW target" curve, in fucsia, isn't *necessarily* my target, depending =
on what
comes from this discussion. It's merely the best fit, per a linear =
regression analysis.</FONT></DIV>
<DIV><FONT face=Arial size=2></FONT> </DIV>
<DIV><FONT face=Arial size=2>Now that I look at this curve a bit =
more -- and
look a bit more at the hammers themselves -- it would appear that there =
is a
natural curve in the set, that is stepped apart across 20/21 because of =
molding
differences. However, the treble end (where the piano is gonna be =
the most
finicky) is straight as an arrow.</FONT></DIV>
<DIV><FONT face=Arial size=2></FONT> </DIV>
<DIV><FONT face=Arial size=2>My original intent was to even out the =
larger jags
in the curve and then to see how the hammers perform in the piano, =
before doing
any more major adjusting. This is in part because my =
keyframe is
vacationing in Florida at the moment, and so this was something I could =
do in
the mean time! <grin> I'm now thinking it would be a =
good idea
to wait until I can install these hammers in the piano and assess their =
initial
performance before doing *anything* to them. Well, it might prove
informative to experiment with the low tenor by wrapping copper wire =
around the
tips of the shanks to bring those hammers up to the linear curve. =
The
beauty of a spreadsheet is that it can tell me what length of copper =
wire to cut
and wrap around each shank. :-)</FONT></DIV>
<DIV><FONT face=Arial size=2></FONT> </DIV>
<DIV><FONT face=Arial size=2>Incidentally, I've thought about how to =
do the
final adjustments. I've of course generated an errors column that =
tells me
how much needs to be added or subtracted from the SW. =
Provided
I'm adding or subtracting weight from the molding itself, this value is
identical to the change in weight of the entire hammer assembly.
Therefore, this process can most easily be performed by weighing the =
hammer (and
shank and flange -- higher than the SW, of course), and adjusting the =
*total*
weight by the same amount. This wouldn't involve the more =
laborious
process of positioning the hammer "just so," with the shank level and
the flange on a stationary support, sticking up exactly 90 =
deg.
Instead, I can simply put the entire hammer on the tray. =
Easy. This
process is made easier if I simply tare the balance to the old weight, =
but I
still need to record the original weight, just in case time runs out, =
and my
balance shuts down. ;-)</FONT></DIV>
<DIV><FONT face=Arial size=2></FONT> </DIV>
<DIV><FONT face=Arial size=2>> > So is this something that is =
the way it
is just because of tradition <BR>> > -- because the cauls are =
built that
way, and that's what ya' get?<BR>> <BR>> It seems reasonable to =
assume
that the chosen manufactured shape of a <BR>> set of hammers would =
directly
affect the basic shape of the curve. <BR>> Cutting of hammer =
sets into
individual hammers no doubt accounts for <BR>> some of the spikes... =
as does
variations in wood density... and probably <BR>> a lot of other =
things I
havent really thought about.<BR></FONT></DIV>
<DIV><FONT face=Arial size=2>I guess I'm thinking more of the linear =
proportions
in the dimensions of the heads. If I'm not mistaken, you =
could take a
slab of uncut hammers and sight a straight line along all =
surfaces.
Correct? (Ray??) Interestingly, the entire keyframe is laid =
out in
straight lines as well, with longer dimensions in the bass and shorter =
in the
treble. Ray explains that the felt in the bass hammers is less =
dense than
in the treble, so that would explain the basic form of the SW =
curve. With
linear dimensions all around, and with progressively decreasing felt =
density,
combined with increasing proportions of felt with the larger/lower =
hammers,
there would actually be a fall-off in the bass end SW.</FONT></DIV>
<DIV><FONT face=Arial size=2></FONT> </DIV>
<DIV><FONT face=Arial size=2>I would think that if hammers were to =
be matched in
mass to the strings they're hitting, the hammers in the high tenor and =
low
treble would be much lighter. But then again, consider the entire =
range of
the piano. It's HUGE. There's a LOT of difference between =
the mass
of the A0 and C8 strings, and this difference is so large that =
matching
hammer mass would be an entirely impractical venture. We'd be =
hitting C8
with paperclips and A0 with bowling balls (exaggerating, of =
course). This
of course leads me back to the question of whether the shape of the =
curve really
means anything. In consideration of the range of the piano, a =
concave
downwards, Stanwood-type curve isn't really much better or worse-matched =
to
reality than a linear relationship. So perhaps the best approach =
is to
simply even out what comes out of the box and work from there.
Hmmmm?</FONT></DIV>
<DIV><FONT face=Arial size=2></FONT> </DIV>
<DIV><FONT face=Arial size=2>I indeed wanted between a Stanwood #8 =
and #9, and
that's pretty much what I got, throughout the bulk of the piano's =
range.
I'm not going to get too worked up if I'm in the #10 range in the high
treble. After all, those notes are pretty easy to play =
anyway!
;-) I think it's all going to come down to tone. Whatever =
needs
doing, that will dictate the shape of the curve (e.g. if some felt =
removal is
ultimately needed here or there, *BUT ONLY WITH THE APPROVAL AND
GUIDANCE OF MY MASTER-SAN*). My job will be to make =
it
smooth. Taking Isaac's advice about touch and tone
being psychoacoustically/perceptually linked (which I highly =
suspect is the
case), I imagine I'll be doing a bit of back-and-forth between touch and =
tone,
until I reach the optimum via the process of successive
approximation. However, I'm going to keep this work to a minimum =
until
AFTER I rescale/restring! </FONT></DIV>
<DIV><FONT face=Arial size=2></FONT> </DIV>
<DIV><FONT face=Arial size=2>Meanwhile, I would still love to know =
if there's
any magic to the shape of these Stanwood curves (David??) or whether =
they
largely reflect the natural curves of hammer sets, resulting from the =
way they
are manufactured (as I suspect).</FONT></DIV>
<DIV><FONT face=Arial size=2></FONT> </DIV>
<DIV><FONT face=Arial size=2>Thanks for the input, y'all! =
Interesting
stuff, to be sure! :-)</FONT></DIV>
<DIV><FONT face=Arial size=2></FONT> </DIV>
<DIV><FONT face=Arial size=2>Peace,</FONT></DIV>
<DIV><FONT face=Arial size=2>Sarah</FONT></DIV>
<DIV><FONT face=Arial size=2> </DIV></FONT></BODY></HTML>