Cool, thank you for the link. To measure and correct for the frequency response of the microphone, can't you just play a range of known frequencies that each have known power levels, and then measure the gain/loss in power level that gets recorded by the mic, and then solve for a filter that gives you unity gain on the mic? I assume that you'd have peaks in the frequency response around the natural resonance and harmonics of the mic, and you'd basically just be adjusting for those.