Action dynamic model

[Cy Shuster]

This is a multi-part message in MIME format.

---------------------- multipart/alternative attachment
It would be difficult to come up with a complete dynamic model -- one =
that
would predict behavior, given a set of inputs.

However it would be easier, and a necessary first step to the =
development of
the model, to measure this finger pressure during the entire keystroke, =
as
Phil describes below.  Graphed over time, it should show lost motion, =
hammer
movement, escape, and aftertouch.  Add the force of the hammer on the
string, and you've got the complete envelope of the mechanical inputs =
and
outputs of the action (well, save for the dampers).

Car magazines publish graphs of car engine performance; wouldn't it be
useful to have graphs like this for a range of test blows from ppp to =
fff?
They could be used to classify actions into categories like "light" to
"firm", as was suggested.

Piano manufacturers have robotic players to settle the action which =
could be
programmed for a series of test blows, as well as the various player
mechanisms (although the former might better simulate a real finger =
pressing
a key).  Do manufacturers take dynamic measurements as well?

--Cy Shuster--
Rochester, MN

>On Friday, May 2, 2003, at 12:48 PM, Phillip Ford wrote:

> As I see it this will be one of the problems that Stephen Birkett (or
> any other contenders) faces in making a dynamic model of an action.
> What input are you going to use?  Does the player push down with
> constant force?  With constant velocity?  With constant acceleration?
> Does he push down with more force initially to break the static
> friction and then back off to a lower force? Does he attempt to
> rapidly accelerate the key in the first part of the movement and then
> maintain a constant velocity for the rest of the
> movement?




---------------------- multipart/alternative attachment
An HTML attachment was scrubbed...
URL: https://www.moypiano.com/ptg/pianotech.php/attachments/84/e4/bc/5d/attachment.htm

---------------------- multipart/alternative attachment--