Loudness and Pitch of Emotional Stage Speech in Kunqu Opera.

Equivalent sound level (Leq), sound pressure level (SPL) and fundamental frequency (F0) were analyzed in the stage speech of six Kunqu Opera roles, Young woman, Old woman, Young man, Old man, Colorful face and Clown. The roles differ in gender, age, personality and phonation types. Differences among emotions (neutral, sad, angry and happy), singers and roles were examined. For most roles, more similarities were observed between neutral and sad stage speech and between angry and happy stage speeches. In most cases, the latter group showed higher Leq, mean SPL and Mean F0 and larger standard deviation (SD) of SPL difference than the former. Some parameters, such as SD of SPL, the mean of SPL difference and the difference between Leq and mean SPL, also showed the intra-group differences.Young woman role, Young man role and Old woman role were similar in some parameters. Colorful face role and Old man role showed a lot of similarities. Clown role showed the least similarities with the other roles. With regard to gender and age, young roles showed smaller SPL difference, larger correlation coefficient between F0 and SPL and larger mean SPL differences between emotions than old roles; female roles had greater parameter consistency and larger correlation coefficient between F0 and SPL than male roles. The personality and phonation types also effected the characteristics of loudness and pitch. This study showed the importance of speakers' characteristics in emotional speech analysis.