The top and back of a guitar, prior to any bracing being applied are frequently referred to as plates. How thick a top plate should be and how the thickness should be distributed in a top plate (in particular) has often been treated as one of life’s great mysteries. Bracing schemes are attributed with similar mystique, though it is readily apparent from the numerous variants in existence that any scheme that is structurally and acoustically efficient will work.
Of critical importance is how to achieve the target vibrational performance of a guitar panel consistently out of the wood that is available. If the plate characteristics can be maintained as a constant and the applied bracing is maintained as a constant the assembled soundboard will have constant characteristics provided there aren’t any radical variations in how it is all put together. We know that the material properties of wood can vary quite substantially. Any two samples of, say, good Sitka spruce will have different densities and long grain and cross grain stiffnesses. So thicknessing these two samples to a constant dimension is guaranteed to yield different vibrational results. Trying to find identical samples is, frankly, a fool’s errand. The trick is in finding a method that will give a consistent vibrational performance out of any reasonable piece of wood. The “tap it and listen” technique of the old master builders is certainly valid for those practiced enough to be able to master it. The technique I use is an engineering version of that. What I do is take a piece of top or back wood, the piece I am about to build with, clean it up to uniform dimensions, analyse the tap tone with a computer and get a figure delivered for how thick to leave it. The derivation of the technique is mathematically involved, but ultimately yields a simple and quick methodology which delivers very consistent results.