We aim to understand the guitar body oscillations, those that provide the "depth" of sound in the lower frequencies, even which musical key an instrument sounds best in. "Tonal ontology", where the tone come from. Here we present our data on arched-plate modal properties, and ultimately guitar modeling using those properties. We want to link the measured plate properties to the body modes of the finished instrument, to guide us in our design and building new guitars.

Tonal ontology part 1: Archtop guitar-plate parameters.

This is a part of tap tuning (but we still don't know what we are aiming for!). There were a few surprises along the way, concerning mode structures and humidity. The next step is to model the guitar body modes using the plate parameters we measured, and see if we can understand what we get.

Tonal ontology part 2: Measured and predicted body modes.

This is where we describe our attempts to predict the box-only and completed-instrument body modes. We found good agreement between modeling and box-only (no neck) frequencies, and did not need to invoke sidewall recoil to obtain a match. This in itself is interesting: Is the agreement fortuitous, is the sidewall modeling incomplete, or are both correct and our supporting system stops the recoil? We should extend experiments to use a different support system and see what happens. f₅ is a cavity (or room) mode and is not part of our modeling.

On the completed instrument the addition of a neck brings in a new resonance. Either the lowest frequency f₁ is split into f₁ and f_1a, or f_1a is new mode. The third highest frequency mode f₃ is either decreased in amplitude and frequency, or disappears, or both: We need more data. There is very limited evidence that it changed over time. Is this because the first lower-frequency isolated and clamped back plate mode disappeared? The modeling shows that suppressing that isolated clamped back plate lower frequency mode would remove f₃, but if this is the answer, why has it happened? Could it be part of the “ageing process”, e.g. a result of our 60 Hz (and harmonics) vibration treatment? Or is f₃ somehow affected by the neck? Again if so, how?

The lowest frequency f₁ is well known to be very dependent on the Helmholtz frequency, and that is hard to measure. It is also dependent on the isolated plate resonant frequencies. These plate frequencies, and the underlying stiffness (we are unsure about the mass), are sensitive to how well clamped the isolated plate is. Clamping off the instrument may impose a different boundary condition from the plate as part of the box. We should experiment with adding mass to obtain modal mass and stiffness not just on the isolated clamped plates, but also on the completed box, and completed instrument. These experiments will be much harder to perform because we must consider all the couplings, and/or invent a better way to immobilize chosen plates.