<html> <head> <META HTTP-EQUIV="Content-Type" CONTENT="text/html; charset=us-ascii"> <meta name=Generator content="Microsoft Word 10 (filtered)"> <style>  </style> </head> <body lang=EN-US link=blue vlink=purple> <div class=Section1> Tensions in a typical well laid out scale will vary about 4-6 lbs within each string gauge section.  Presetting a completely smooth tension curve as a starting point will give you problems.  Most good scales are laid out in logarithmic progressions which allow you the smoothest and most regular transitions between gauges, keep tensions reasonably smooth (at least as far as the trend line goes) and produce a smooth inharmonicity curve.  On a given scale, i.e. original bridge) you will have better luck filling in the gauges by hand observing what happens to the tensions and inharmonicity (not to mention impedance and break point percentage).  The problem with more poorly laid out or original scales will be where to compromise between tension and inharmonicity (the two most important factors in my view) while keeping an eye on other factors.  There are differences of opinion on what the most important factors to consider are and when analyzing some scalings, even current ones, you see some pretty weird stuff.  Not likely the program will be able to do that automatically as setting up guidelines that take into account the relative weighing of each factor will be difficult, if not impossible.  There is art as well as science to this part.  You need to first decide on which part of the science you want use and then start considering the art.  Limitations of string winding (available diameters of core and wrap wire) must also be considered.  You really need to do much of it manually.    Good luck.     David Love davidlovepianos@comcast.net www.davidlovepianos.com   RicB wrote:   > Hi folks > > As you know Jason and a few friends are working on a spreadsheet for > designing scales to share with the entire list.  He / we are trying to > include a few specialized automation features and one of them is > proving a bit challenging.  The idea is to try to provide some kind of > way of automatically evening out tension as much as possible while > maintaining a resonable string length curve and string diameters.  > Interestingly... the first attempts which floated diameters only when > tension was all alligneed to an exact value....  resulted in a really > weird (by usual standards) set of string diameters... with hops up and > down in size all over the place.  Inharmonicity was not looked at (yet). > > Questions that arise have to do with what kind of range for variance > in tension to scale designers typically operate with, and ... then I > was wondering what really would be so bad about a scale that had > diameters not necessarilly decreasing in size evenly up the scale... > what would be wrong with a more ... jagged curve ? > > Jason is putting a lot of effort into this so any feedback would be > greatly appreciated.  Any old scales you have stored for studying > should be sent along to him. > > Cheers > RicB > >   </div> </body> </html>