Abstract: This paper introduces a novel predictable voice morphing system. It is superior due first to the use of the STRAIGHT model that allows flexible manipulation of speech parameters such as pitch, vocal tract length, and speaking rate while maintaining high reproduction quality. The advantage of the system is also attributed to the introduction of the predictable spectrogram, resolving the problems of over smoothing of GMM mapping, and discontinuities between consecutive frames caused by traditional LPC model. Subjective evaluation and objective measurement indicate that the proposed method outperforms the traditional method both in synthesized quality and precision of mapping target characteristics.

Key words: STRAIGHT model, predictable pitch, predictable spectrogram, voice morphing