Overlap and Add Sinusoidal Synthesis Method of Speech Signal using Amplitude-weighted Phase Error Function and Damping Harmonic Amplitude Parameter

In this paper, we propose new methods with the improved continuity performance of over and add speech signal synthesis method using amplitude-weighted phase error function and damping harmonic amplitude parameter. The general overlap and add sinusoidal synthesis method minimizes the phase discontinuity of the boundaries of the sinusoidal synthesis signal to estimate a mid-point phase value. But, one of the proposed methods minimizes the wave discontinuity of the synthesis signal to estimate the mid-point phase. The proposed method uses the average value of past and current parameters for the sinusoidal amplitude used as the weight of phase error function. But, another proposed method extracts the more accurate sinusoidal amplitude by using a correlation between the original signals and the synthesized signals for the sinusoidal amplitude used as the weights. To verify the performance of the proposed methods, we observed the average differential error value between the synthesized signals.