Abstract
Acoustic Analysis and synthesis of hoarseness by sound spectrography suggests that the acoustic properties of hoarseness are mainly determined by the interactions of the following three factors: 1) noise components in the main formant of each vowel, 2) high frequency noise components above 3000 Hz, and 3) the loss of high frequency harmonic components. These three findings are more pronounced in the vowels /α/, /ε/, and /i/ than in /u/ and /ɔ/. With the progression of the severity of hoarseness, these three abnormal patterns become more prominent and exaggerated. On the basis of these findings, a classification of four types of hoarseness was presented using sonagram tracings.