[CAUT] Soundboard weighting

[Ron Nossaman]

Greg Newell wrote:

> 
>                 I’m confused. If the purpose of the treble fish is to 
> add stiffness why would you want to thin the board behind the bridge and 
> before the fish? Is that negating the effect of the fish? I thought that 
> the whole idea of the fish was to make a tighter system up there? Would 
> thinning loosen it? Isn’t the purpose of the mass weighting to add even 
> more tightness / stiffness/ lack of flexibility?

Greg, David,
Come on guys, we've gone over this a lot of times. There are 
two soundboard considerations here, impedance, and resonant 
frequency. From an impedance standpoint, the board needs to be 
stiff enough to not just absorb string energy like a loose 
awning (or killer octave), producing the classic high treble 
"dink", with no sustain. From a resonant frequency standpoint, 
if the board is too stiff and light, the response frequency 
overruns the driver (string) and the thing shrieks into chaos.

Go to: 
http://www.kettering.edu/~drussell/Demos/SHO/mass-force.html

This is a demonstration of the resonant frequency thing.

The first example is representative of a treble that is too 
stiff and light, having a resonant frequency higher than the 
driver. The response is very disordered, which technically 
qualifies as "ungood".

The second example, though that wasn't it's intent, does a 
pretty good impression of a board that's not stiff enough.

The third example is like a board that is of a proper mix of 
stiffness and mass as to have a resonant frequency lower than 
the driver, so it follows along docilely like it should.

When building production pianos, you have the luxury of making 
a few, diddling parameters until you find something that works 
well enough to crank 'em out like that in perpetuity. In a 
rebuilder's shop, you're doing a one off, so you only get one 
guess. Compounding that is the attitude that, since you've 
deviated from the original design, just making it better than 
it ever was, or better than all the others like it ever were, 
isn't enough. The damned thing has to be PERFECT, and exceed 
the most minute details of the wildest dreams of anyone who 
even hears rumors of it's existence until you build the next 
one, which is expected to be that much better. So you can't do 
a whole lot of random guessing. You apply principles you know 
work with some predictability. So you either shoot for the 
middle of what you think will prove to be adequately stiff, 
and hope for the best, or you build what you know is a bit too 
stiff, leaving a safety net of added mass to fall back on. 
Your call. I, personally, prefer to build in some tolerance, 
and a second chance, where I can.
Ron N