[SpruceTop]

From what I can gather, mic placement for training the ToneDexter is crucial to generating the tone a player wants to hear in their amplified sound. Maybe for those wanting a fuller amplified tone, the mic should be aimed between the neck/body juncture and the soundhole of the guitar, rather than the more generally-accepted neck/body juncture? Or, how about aimed at the soundhole but from an experimentally determined distance to minimize boominess while retaining fullness?

Then there are the characteristics of the mic itself that should be taken into consideration to help expedite the training process, For example, if I were to use an Ear Trumpet Labs Edwina for training the ToneDexter, the company's Edwina analysis graph shows the Edwina at:

6 inches from sound source as having a Flat frequency response 20Hz - 15kHz (-3dB) with 80Hz (-1dB);

and

20 inches from sound source as having a frequency response of 20Hz (-7dB) - 15kHz (-3dB) with 80Hz (-6dB)

So, for training the ToneDexter to have a neutral or flat EQ frequency response with regards to the Edwina's characteristics, I'd start by placing the Edwina about 6 inches from the guitar, and aimed at several different places on the guitar to generate several different ToneDexter files and choose the best result. Then, I'd move the mic back a few inches and do the same training procedure and choose the best result. Of course, depending on the guitar, there will be boomy- and thin-toned results at various distances and aiming directions but at least there will be an organized procedure to find the most pleasing Dexterized tone.

Trumpet Labs Edwina Analysis Graphs