Influence of sampling rate on accuracy and reliability of acoustic voice analysis.
Influence of sampling rate on accuracy and reliability of acoustic voice analysis.
复制标题
DOI：
10.1080/1401543051006721
复制
发表时间：
2005-01-01
期刊：
影响因子：
--
通讯作者：
Evans, Maegan K
中科院分区：
文献类型：
--
作者：
Deliyski, Dimitar D;Shaw, Heather S;Evans, Maegan K
众所周知，采样率(F(S))影响声学语音测量的可靠性和有效性，但还没有确定一个确切的关系。本实验的目的是考察F(S)对声学语音质量测量的影响，同时考虑性别、受试者内部变异性、麦克风、环境噪声、数据采集硬件和分析软件的影响作为平衡因素。通过使用MDVP、TF32和PRAAT三个软件分析系统，通过分析864,000个基频、抖动和微扰，探讨了从44.1千赫兹到10千赫兹的F(S)的影响。结果表明，声学语音分析的推荐、可接受和临界F(S)分别为26 kHz以上、19 kHz以上和12 kHz以上。因此，捕获的26 kHz以上的语音样本可以用于数据分析和比较，而不会引入F(S)带来的误差。
It is universally recognized that sampling rate (F(S)) influences the reliability and validity of acoustic voice measurements; however, an exact relationship has not been determined. The purpose of this experiment was to investigate the influence of F(S) on acoustic voice quality measurements, while considering the influences of gender, intra-subject variability, microphone, environmental noise, data acquisition hardware, and analysis software as balancing factors. The impact of F(S), from 44.1 kHz to 10 kHz, was explored by analyzing 864,000 measures of fundamental frequency, jitter, and shimmer, using three software analysis systems: MDVP, TF32, and PRAAT. Results suggest that the recommended, acceptable, and critical F(S) for acoustic voice analysis are above 26 kHz, above 19 kHz, and 12 kHz, respectively. Thus, voice samples captured above 26 kHz can be used for data analysis and compared without introducing error due to F(S).