Hammers and scale

[Ron Nossaman]

>I guess I'm talking about an "all other things being equal" situation.  If,
>on a given piano, one determines that the tensions in the treble needs to be
>increased, for example, to bring the tension curve more in line with the
>rest of the piano, would you expect to have to reconsider the hammer style
>to accommodate that change.

No, I wouldn't.


>To put it a different way, I notice that the tensions on some Steinways tend
>to be a bit low crossing the treble break.  Many formulas for calculating
>the scale call for an increase in the wire diameters in those sections.

You see this in a lot of pianos. It's still being "designed" into new ones. 
Insufficient bridge dog leg at the plate struts in the treble disrupt the 
speaking length log progression. The fix is simple. Replace the bridge. The 
low tensions in the low tenor are a result of the foreshortened speaking 
lengths the hockey stick back curve of the bridge produces. I'll never 
understand the insistence of putting a 20 note bass in a scale that needs a 
28, 30, or 32 note bass. The fix is to go back in time and design the piano 
with a realistic number of notes in the bass section, placing the break 
farther up scale and allowing for a full length low tenor that is at a 
reasonable tension. The next best alternative is to add a transition 
bridge, bringing the bass/tenor break up into the tenor. Lacking that, the 
next best hope is adding some wound strings to the low tenor and trying to 
blend everything as best you can - which probably won't be as good as you'd 
like, but will at least be a different set of problems than was there 
originally and it will probably sound better than the original bonk bonk. 
The next option would be to just go with a larger wire size and trade bonk 
bonk for thud thud at a higher tension. While it's possible to put a 
terrible scale on a nice set of speaking lengths, you're at the mercy of 
the bridges when you're trying to improve an existing scale. You do what 
you can.

Then there's the soundboard. If soundboard impedance is too low at the low 
tenor because of the rib scale, or the bridge being too short and not 
providing adequate stiffness, there isn't a thing you can do with string 
scaling to fix it. Up among the treble plate struts, you'll find a similar 
low impedance condition that you likewise won't improve with scaling. A 
48mm C-8 on a compression crowned board might dink more definitively with 
0.032" wire than with 0.031", or it might not. But it probably will dink, 
nonetheless. Again, you're at the mercy of the soundboard and bridge.

And this is all before you get anywhere near a hammer.


>What, in your experience, explains the disparity between the original scale
>design and the recalculated scale that calls for that type of change.  Is it
>a difference in general philosophy?  And how would you characterize that
>difference.

I don't know what the general philosophy behind the Steinway scales was, 
but I suspect they would be different if they had been designed with the 
tools we have today.


>Is there something in the Steinway soundboard assembly that
>calls for that type of original scaling?

No, but the soundboard assembly does result in the characteristic "voicing" 
problems we have with them.


>Does putting a higher
>tension scale in that section create other tonal problems?  Would a slightly
>higher tension cause you to reconsider how you set downbearing, for
>instance?
>
>David Love

Probably not, in both cases. From a structural standpoint, I'd keep 
tensions fairly close to what they were originally, particularly around the 
bass/tenor break.

Ron N