This document discusses methods for running pitch detection on mobile devices. It compares zero crossing, autocorrelation, and harmonic product spectrum (HPS) algorithms, and recommends FFT and phase shift. The suggested process is to window the input signal, run FFT to calculate the phase shift, identify peaks, and store results for a duration. Volume in decibels can indicate voiced and unvoiced sounds. FFT has limitations in range, noise, and accuracy, but can be optimized for real-time use on mobile through accelerator libraries and multithreading.