I have a working algorithm. have I said today ... damn I'm good. (actually I did multiple times at work ( I kid you not) ).
Now time to optimize the FFT audio transform for peak performance and tweak it so it fits my purpose rather than its current implementation.