Hey everyone ,i've got problems , i can use FFT algorithm to draw signal ( amplitude change with time) , and from there i can draw frequency response (amplitude (energy) versus frequency) . My teacher said that i can still use it for classifying a audio (is it a speech or a music ? ) but i had no idea ,after that i found down that i can classify them by using the SRs ( silence ratio ) , harmonic and pitch . Then the problems came out , i've got no idea how to calculate it or detect it . So if anyone have any ideas , please show me . I code in C# .