Perkell J S, Hillman R E, Holmberg E B
Research Laboratory of Electronics, Massachusetts Institute of Technology, Cambridge 02139.
J Acoust Soc Am. 1994 Aug;96(2 Pt 1):695-8. doi: 10.1121/1.410307.
In previous reports, aerodynamic and acoustic measures of voice production were presented for groups of normal male and female speakers [Holmberg et al., J. Acoust. Soc. Am. 84, 511-529 (1988); J. Voice 3, 294-305 (1989)] that were used as norms in studies of voice disorders [Hillman et al., J. Speech Hear. Res. 32, 373-392 (1989); J. Voice 4, 52-63 (1990)]. Several of the measures were extracted from glottal airflow waveforms that were derived by inverse filtering a high-time-resolution oral airflow signal. Recently, the methods have been updated and a new study of additional subjects has been conducted. This report presents previous (1988) and current (1993) group mean values of sound pressure level, fundamental frequency, maximum airflow declination rate, ac flow, peak flow, minimum flow, ac-dc ratio, inferred subglottal air pressure, average flow, and glottal resistance. Statistical tests indicate overall group differences and differences for values of several individual parameters between the 1988 and 1993 studies. Some inter-study differences in parameter values may be due to sampling effects and minor methodological differences; however, a comparative test of 1988 and 1993 inverse filtering algorithms shows that some lower 1988 values of maximum flow declination rate were due at least in part to excessive low-pass filtering in the 1988 algorithm. The observed differences should have had a negligible influence on the conclusions of our studies of voice disorders.
在之前的报告中，针对正常男性和女性说话者群体给出了语音产生的空气动力学和声学测量结果[霍尔姆伯格等人，《美国声学学会杂志》84卷，511 - 529页（1988年）；《嗓音杂志》3卷，294 - 305页（1989年）]，这些结果在嗓音障碍研究中用作标准[希尔曼等人，《言语听觉研究杂志》32卷，373 - 392页（1989年）；《嗓音杂志》4卷，52 - 63页（1990年）]。其中一些测量结果是从通过对高时间分辨率的口腔气流信号进行逆滤波得到的声门气流波形中提取的。最近，这些方法已经更新，并对更多受试者进行了一项新的研究。本报告给出了1988年（之前）和1993年（当前）的声压级、基频、最大气流下降率、交流气流、峰值气流、最小气流、交流 - 直流比率、推断的声门下气压、平均气流和声门阻力的群体平均值。统计检验表明1988年和199年研究之间总体群体存在差异以及几个个体参数值存在差异。参数值的一些研究间差异可能是由于抽样效应和微小的方法学差异；然而，对1988年和1993年逆滤波算法的比较测试表明，1988年最大气流下降率的一些较低值至少部分是由于1988年算法中过度的低通滤波。观察到的差异对我们嗓音障碍研究的结论应该只有可忽略不计的影响。
