A method for processing an audio signal (i(t)), comprises: receiving a first set (x(t)) of time-varying signals representing a first sound comprised in the audio signal (i(t)), the first set (x(t)) of time-varying signals comprising an amplitude modulation signal (a(t)), a carrier frequency signal (fc(t)), a pitch signal (f0(t)) and an FM index signal (h(t)); and modifying the first set (x(t)) of time-varying signals by modifying the amplitude of the FM index signal (h(t)), thereby providing a first modified set (x′(t)) of time-varying signals. The resulting first modified set (x′(t)) of time-varying signals may subsequently be modulated to provide an audio output signal.