Patent claims priority to a July 5, 2013 provisional and covers detecting emotion from a user's voice and generating adaptive speech based on pitch and harmonics