METHOD FOR ENHANCING SOUND QUALITY OF VOICE VOCODER THROUGH FORMANT POST-FILTERING USING MULTI-ORDER LPC

摘要

PURPOSE: A method for enhancing sound quality of a voice vocoder through formant post-filtering using multi-order LPC(Linear Predictive Coefficient) is provided to remarkably enhance the sound quality using multi-order compensation filters such as second-order or third-order compensation filters instead of a first-order slope compensation filter. CONSTITUTION: First, a self correlation coefficient having delay as much as desired order number is extracted. Then, an energy value for a current sub frame is extracted. Next, a self correlation coefficient according to the ratio of the two values is calculated. Then, a new self correlation coefficient is generated by being combined with a self correlation coefficient used in a previous frame for obtaining a final self correlation coefficient to be used in a filter. Finally, a slope compensation filter is structured by using the coefficients.