// TODO: float is "sqrt(1 - coherence)". Here MAX_16 is used for 1. coherence_fx is in Q15. Why isn't ONE_IN_Q15 used here? (which is different from MAX_16)
// TODO: float is "sqrt(1 - coherence)". Here MAX_16 is used for 1. coherence_fx is in Q15. Why isn't ONE_IN_Q15 used here? (which is different from MAX_16)