Thanks for the support Micheal. I've been busy for a couple of days, and also pondering a response, and that input helps.
I think you have to admit that your screen name is not very informative. It says nothing to me about your experience, and leaves little to go on other than the posts I read. Anyboody who knows me at all will attest that I'm not a confrontational person, and it was never my intent to call your experience into question; I simply had no way of knowing what it was.
Until I started measuring the properties of the tops I got, I really didn't know how variable they were, nor how little information you actually get visually. This is really all I've been trying to get across.
A collegue of mine took summer courses in violin making at the University of New Hampshire that were taught by the head master of the Mittenwald school, Karl Roy. Roy used to gauge stiffness by flexing in various ways, and one of the students wondered how accurate he was, so he made up a box full of small sample pieces of various sizes and measured them all. He then asked Roy to sort them out by stiffness, doing the test several times during a one-week session. He found that Roy could consistently distinguish a difference of about 3%, which is about as well as you're going to do with a good 'shop level' test. So it is possible. What's important in that is the Roy did not determine the thickness based on appearance, but on measurement that he made using his hands, and based on many years of experience. I must assume for purposes of discussion that your skill appproaches that level, and have no doubt that you can properly gauge the thickness of a top you are working on. What I find difficult to credit, based on my own extensive experience, is that anybody could properly gauge the thickness of a top on appearance alone.
"Do you mean by stiffness "Young´s modulus"?
Does Young´s Modulus directly linear correlate to the specific weight/density, i.e., the heavier the stiffer?"
Yes to both questions, in a very general sense. We are, after all, talking about a natural material here, so you can't expect perfect uniformity. As it is, in using softwoods for tops we are reducing the variation by quite a lot; as it happens all softwoods have a very similar microscopic structure, so there is far less variation across softwood species than there is with hardwoods.
I don't pretend in the following to tell you anything you don't probably already know, but I'll include the explanation for clarity for those who don't.
The stiffness of a piece of material of a give size is largely determined by it's Young's modulus, which is a measure of the work required to stretch or compress it. Bending something stretches and compresses the material, especially near the surfaces. Thus the stiffness depends on the size, and particularly the thickness/height of the piece, and the Young's modulus. All else equal, stiffness goes as the cube of thickness/height, so small differences in thickness can change the stiffness a lot: 25% thicker will be almost twice as stiff, for example (but only 25% heavier, of course). This is why it is improper in a strict sense to talk about the 'stiffness to weight ratio' of a material
: you need to specify the section for that. That cubic realtionship also is one of the main problems in getting accurate measurements of the Young's modulus; to get within 3% of the modulus value you need to control the thickness within 1/10% or so, which is hard when you're down in the range of normal guitar top thicknesses.
Plots I've made of the reationship of Young's modulus along the grain of various top woods to density show most samples falling close to the same line. Within the 'normal' range of densities; say from roughly 300-550 kg.m^3, it's very nearly a linear relationship. 2/3 of my samples fall within 10% plus or minus of that line, which is quite good considering normal measurement errors. These values are quite repeatable, again, given the influece of things like humidity changes.
Late wood tends to contribute more to density than Young's modulus; the outliers that are stiffer than expected tend to have relatively narrow latewoood lines. Heavy latewood, and particularly 'reaction (compression) wood' goes the other way. Note that this is not grain count
, but the relative width of the early and late wood. So there's one thing that you can see that is a rough indicator of stiffness to density. Another is run out, which reduces the Young's modulus, as you'd expect.
Note that all
of the softwoods I've tested fall on the same line for lengthwise Young's modulus vs density. That includes 'usual suspects' like Western Red Cedar, Engelmann spruce, European spruce, Sitka spruce, Red spruce, and Redwood, as well as soome things like White pine, Western Hemlock, White spruce, and some others. Other properties vary; both Redwood and WRC have much lower damping than the spruces in general, for example, but in terms of Young's modulus and density they follow the same rule.
There is a lot of variation in density within any species. Although, for example, Western Red cedar tends to be lower in density on average (in my samples so far) than the spruces, it's not at all hard to find relatively dense WRC, or fairly light Sitka spruce.
Cross grain Young's modulus in softwoods primarily relates to how well quartered the piece is. Even a small deviation from 'perfect' quarter can give a large drop off in cross stiffness. The 'best' crosswise Young's modulus value for different trees of the same species varies somewhat, but it's hard to get enough data to sort out the variables there.
I'm presently of the opinion that having a top with the highest possible cross grain stiffness may not always be necessary, or even desireable. There are reasons to question whether it contributes much to the overall stiffness of the top over the long run, as 'cold creep' (top bellying) seems to cancel it out to some extent, at least as a structural attribute. Acoustically it's at least plausible that there is a 'best' cross grain stiffness for a given outline of top. Again, I'll say that this in my cuurrent thinking on this, and it's working pretty well, but I do reserve the right to change my mind if better data comes in.
Note that all of this data comes from measurements of top half blanks. We all know how much a piece of wood can vary from one spot to another, and these measurements smear all of that variation out. That variation could have a lot to do with why a good maker would vary the thickness of a top from point to point.
So there's a long answer to your short questions.
I'll note how little of this has anything to do with the usual visual markers of 'quality' in a top, such as tightness and straightness of grain, uniformity of color and so on. In that respect it's entirely possible to have a top that is of 'high quality' but is still not as good acoustically
as one that is of a lower grade. I have paid top dollar for 'high quality' tops that I considered barely usable when I recieved them, and gotten very good tops that were very low in visual grade.