共振峰编辑法区别鼻化元音中口、鼻音共振峰的实证探究*

为解决司法话者识别中利用鼻化元音构建元音声学空间图时如何准确判别鼻化元音的口、鼻音共振峰的问题。本文通过计算机语音工作站对语音样本的共振峰进行编辑操作,利用生成的语音样本构建不同的对照组分别进行听辨。结果表明,口音、鼻音共振峰分别被衰减后的语音变化特点呈现一定规律,使用此方法可以准确区分鼻化元音的口、鼻共振峰的阶次。本文建立的“共振峰编辑”与“听觉感知”相结合的判别方法,可以为司法话者识别及语音感知、识别等相关领域通过构建元音声学空间图进行声学特征研究的模型提供口音、鼻音共振峰的判别依据。

An empirical study of the formant editing method to distinguish the oral and nasal formants in nasalized vowels

In order to solve the problem of how to accurately distinguish the oral and nasal formants of nasalized vowels when using nasalized vowels to construct the acoustic space map of nasalized vowels in forensic speaker comparison. This article uses the computer voice workstation to perform formant editing operations on the voice samples, and use the generated voice samples to construct different reference groups for listening identification separately. The results show that the voice change characteristics after the oral and nasal formants are attenuated respectively show certain rules. This method can accurately distinguish the order of the oral and nasal formants of the nasalized vowels. The method of combining "formant editing" and "auditory perception" established in this paper can provide a basis of the identification of oral and nasal formants for the models, which conducts acoustic research by constructing an acoustic space map of vowels in forensic speaker comparison, speech perception, recognition and other related fields.