Virtual sound generation by linear modal synthesis based on recorded audio examples