To improve the clearness degree of a voice signal, to which speaking speed change processing is performed, by providing a consonant amplifying means for amplifying a consonant part preceding to or/and following a vowel part extended by an extending means.

The voice signal fetched by a microphone 2 is inputted through an amplifier 3 and a filter 4 to an A/D converter 5 and converted to a digital signal by the A/D converter 5. Then, the voice signal converted to the digital signal by the A/D converter 5 is fetched and processed. Namely, for the inputted voice signal the vowel part is extended and the consonant signal is amplified. First of all, concentrating the vowel part, the vowel parts in three syllables from the top of inputted series of voice signals are extended. Next, the consonant parts preceding to and following the extended vowel parts are amplified. In this case, the consonant parts in the 4th syllable from the top are amplified and the amplification is inhibited for the following consonant parts. Thus, the voice signal is prevented from getting unclear.