首页|基于EGG的藏语安多方言塞音发声研究

基于EGG的藏语安多方言塞音发声研究

扫码查看
借助电声门图采集了藏语安多方言塞音的生理语音数据,根据嗓音发声的特征矩阵,对3类塞音的发声类型进行了研究。通过混合算法求得开商和基频等生理参数,并对不同发音部位的塞音进行了平滑样条方差分析。实验结果表明,藏语安多方言单辅音中的浊塞音已发生浊音清化现象,变为了同发音部位的清不送气音。从安多方言塞音发声的特征范畴可以看出浊塞音为紧喉嗓音,清送气塞音和清不送气塞音为正常嗓音,开商值可以有效区分这两类嗓音发声,而基频和基频抖动同样也具有参考意义。
A Study on the Phonation of Plosives in the Amdo Tibetan Dialect Based on EGG
This paper collected physiological phonetic data of the Amdo dialect of Tibetan language using electroglottographic method,and studied the phonation types of three types of plosives based on the feature matrix of vocal sound production.The physiological parameters such as open quotient and fundamental frequency were obtained through a hybrid algorithm,and the smooth spline variance analysis was conducted on the physiological parameters of different articulatory positions of plosives.The experimental results show that the voiced plosives in the Amdo dialect of Tibetan language have undergone the process of voicing to devoicing,becoming the same articulatory position's voiceless unaspirated sounds.From the articulatory feature scope of Amdo dialect plosives,voiced plosives are tense vocalic sounds,voiceless aspirated plosives and voiceless unaspirated plosives are normal vocalic sounds.The open quotient value can effectively distinguish these two types of vocalic sounds,and the fundamental frequency and fundamental frequency jitter also have reference value.

Amdo dialectPlosiveElectroglottographtPhonation

完玛吉、更太加、龙代吉草、魏建国

展开 >

青海民族大学国家民委重点人工智能应用技术实验室,西宁 810007

天津大学认知计算与应用重点实验室,天津 300350

安多方言 塞音 电声门图 发声类型

2024

青海科技
青海省科学技术厅

青海科技

影响因子:0.052
ISSN:1005-9393
年,卷(期):2024.31(5)