[CAUT] stack fit to keyframe

[John Delacour]

At 8:09 am -0500 11/9/06, Bob Hohf wrote:

>I'm afraid I can't describe briefly an idea that took a 4-article series to
>develop.  I would at least have to reread the articles to see what I said
>then ...

That is quite understandable, and to respond to all the points you 
raise in your message would produce too long a reply, so I'll limit 
myself to the question of the hammer and the question of the model.

>...I mention having the centerline of the hammer perpendicular to 
>the string when striking.  This is debatable, and in reality is 
>almost never the case.

The only debatable point for me in practice is how much to allow for 
settling-in.  I aim to have every hammer strike the string at a right 
angle once all toning and facing is completed and the piano has been 
well played in, though in practice this adjustment is very slight and 
particularly slight in the extreme treble, where the angle is most 
critical.  I have never yet come across a piano that has a constant 
string height at the strike point, and this is the first variable I 
take into account when calculating how to bore the hammers. All else 
being equal, this almost always means that a longer bore-length would 
be required at note 30 than at note 88.  But all else is not equal, 
since the angle of the strings to the horizontal is generally 
negligeable at note 30 and very considerable at note 88.  It is by 
taking into account both these variables that the hammers are bored 
with the aim a) to respect a straight strike-line b) to have the 
angle (at strike, not necessarily 180 deg.) of the shank to the 
key-bottom the same throughout the scale and c) to have each hammer 
strike the string at a right angle.


>Using an action model modified to make the wippen and hammershank centers
>adjustable, I varied them, graphed it out and determined that optimizing
>this set of parameters specifies a single optimum string height. I suppose
>choosing different parameters might indicate a different height.  I have
>never explored this idea and am unlikely to unless someone wants to finance
>the research.

In the past few weeks I have been working on a script that creates 
drawings from which a QuickTime movie is produced automatically 
showing the full action cycle very precisely and realistically.  As 
it develops I am able to adjust any of the key data and observe the 
results.  As much technical information (arcs, angles, dimensions 
etc.) as is required can be printed to the movie.  As it is at the 
moment it is very impressive, but in fact there remains a lot of work 
to do since I have in places substituted plausible fudges for strict 
trigonometry.  The next main task is to discipline my brain to get 
the angles precise, and after that to introduce a variety of 
different lever designs as options.  At the moment I am using a 
Herrburger lever of about 1898, which I consider to be the best 
development of the Erard-Herz principle ever put into production, but 
I intend soon to add the more common 1850s style lever as perpetuated 
by Steinway and other fossilised companies.

When the script is further advanced I hope to make the movie(s) 
available and eventually perhaps to allow users to send data to a 
server and have a movie produced according to their own 
specifications.  At any rate I shall be producing a form for data 
entry from which the script will be able to create customised 
drawings and movies based on the required measurements.

JD