Abstract
Five reference points were marked on the face of 12 adult subjects in midsagittal plane. The speech sample contained 25/s/and 25/m/sounds along with vowels and/sakuranohanagasakimashita/.
Three phenomena were simultaneously recorded by tow high speed TV cameras on same video tapes. These were (1) the motion of facial reference points or the mandibular movements on a MKG, (2) the spectrogram of the voice analyzed by a SSD (Speech Spectrographic Display), (3) the voice of subjects. The record was playbacked on a monitor of high speed TV. The flames, at which consonants were pronounced, were identified by the spectrogram on the SSD. The positions of reference points were computed by the analyzer (Sportias 300).
The results showed that the jaw and the facial reference points moved to different direction, and took specific positions during pronunciation of/s, sh, k, m, h/. In general, the displacement pattern of the jaw was similar to that of the chin. But, in/s, sh/, the movement of the chin was greater than that of the jaw, and in /m/, the chin moved in the opposite direction of the jaw.
