VOICE ENCODING DEVICE AND VOICE DECODING DEVICE

摘要

PURPOSE:To prevent deterioration in sound quality when a voiced sound which has high pitch periodicity is reproduced as to the CELP type voice encoding device including constitution which feeds the composite signal of the periodic component and a periodic component of a sound source back to the generation part for the periodic component of the sound source and generates the periodic component of the sound source. CONSTITUTION:The periodic component and aperiodic component of the best sound source signal with which a reproduced signal which is most approximate to the input voice signal is reproduced are inputted and according to the relative sizes of the periodic component and aperiodic component of the best sound source signal, when the period component is relatively large, a feedback component correcting means 7 suppresses the aperiodic component more and supplies it as an input signal to a periodic component generating means 1 which generates the periodic component of the sound source signal following up the input signal.