Hi Stephen, > I'm thinking something is causing this that > occurs before the string impact. The obvious candidate is interaction > between the backcheck and the hammer tail as the hammer heads toward > the string. Backcheck clearance is supposed to be pretty tight, and > I've seen enough examples of slight catching on the check as the > tail goes up, enough that power is lost without it being obvious. What a compelling theory! I think you just may have something there. We are to understand that with forceful play, it's possible to bottom a key before the hammer even budges (er... *much*). However, with the bottoming of the key, the backcheck would be elevated and could obviously catch the hammer before it can accelerate out of the way -- if the clearances aren't adequate. I would presume that the only thing that would prevent this is the flexure of the various action parts. The keystick would be flexed concave-downwards, such that the backcheck would be bowed out and downwards (away from the hammer). At the same time, the jack would be punched up into the knuckle, and the hammer shank would be bowed concave-downwards, causing the tail of the hammer to flex away from the backcheck. I'm wondering if all this can't be measured with a "pseudo-dynamic" model. Could the hammer be restrained while the keystick is bottomed out? It seems like that would be kinda stressful to me, but I guess forceful playing is rather stressful. (As a kid, I remember watching Van Cliburn snap a hammer shank in concert, which a technician plucked from below the strings between movements.) Honestly, I'd be a bit afraid to try it on my own piano, without someone I trust reassuring me it's really OK! ;-) If this can be done, then flexure and clearance can be measured in a reasonably controlled manner. This "pseudodynamic" model would of course not take into account the flexure of the keystick between the capstan and the tip or the flexure of the backcheck wire/post (term??), due to acceleration, nor would it account for acceleration in other parts. It would simply account for what are surely the most prominent variables -- hammer flange and keystick flexure (or at least most of it), felt compression, etc. While I realize you probably have the high-speed video equipment and electronic instrumentation to take true dynamic measurements, I'm thinking some sort of pseudodynamic measurement system would be extremely useful to the technician in the field -- since not all hammer/shank assemblies and (especially) keysticks are created equal. I was also thinking about the suggestion of steepening the backcheck angle to create a more positive touch. If indeed the hammer and backcheck are meeting, due to action part flexure, then the problem would certainly be aggrivated by having too steep a backcheck angle. There is probably some angle that is "just right," yielding just enough clearance on a hard power stroke, while preserving a tight check on the rebound. It's interesting that centuries of piano building have resulted in the trial-and-error optimization of action geometry and regulation standards, and it's only recently that we're trying to figure out why these arrangements work. Fascinating stuff! I look forward to hearing what you folks figure out! Peace, Sarah PS Thanks, Robin! I enjoy your insights too. :-)
This PTG archive page provided courtesy of Moy Piano Service, LLC