A New Method for Deconvolution of Steady State Responses