Theory of Speech masking by reverberation
Theory of Speech masking by reverberation
复制标题
混响语音掩蔽理论
DOI：
--
复制
发表时间：
1949
期刊：
影响因子：
--
通讯作者：
A. MacDonald
中科院分区：
文献类型：
--
作者：
R. H. Bolt;A. MacDonald
本文提出了混响掩蔽效应对语音清晰度影响的一般统计理论。语音被认为是一系列离散脉冲，在给定频带的声压级中统计分布在30 dB范围内。清晰度指数计算作为混响时间的函数，使用从可见语音频谱图获得的语音脉冲长度和间距的初步值。然后从清晰度指数计算单词的清晰度百分比，并与努森的实验值进行比较。理论值与混响时间小于2秒的测量值完全一致，在6秒内相差不到17%。计算扩展到包括背景噪声和混响的组合。
A general statistical theory is developed for the masking effect of reverberation on the intelligibility of words. Speech is considered a series of discrete pulses distributed statistically over a 30‐db range in sound pressure level in a given frequency band. The articulation index is calculated as a function of reverberation time, using preliminary values of speech pulse lengths and spacings obtained from Visible Speech spectrograms. The percent articulation for words is then calculated from the articulation index and is compared with Knudsen's experimental values. The theoretical values agree precisely with the measured values at reverberation times less than two seconds and differ by less than 17 percent out to six seconds. The calculations are extended to include a combination of background noise and reverberation.