Proposal of a Japanese-speech-synthesis Method with Dimensional Representation of Emotions based on Prosody as well as Voice-quality Conversion