利用虚拟声环境研究教室语言清晰度

Research on Speech Articulation in Classrooms Under Virtual Listening Environment

摘要:

教室室内语言清晰度是教室声学研究的重点,而语言清晰度试验是研究教室室内听闻环境最直接的方法.但要找到声环境按照一定规律变化的教室并完成语言清晰度试验,往往很困难.处理教室实测脉冲信号,采用指数衰减法可以构建混响时间按一定规律变化的声环境,而采用噪声混入法则可以获得信噪比不同的虚拟声环境.通过对虚拟声环境下的语言清晰度试验结果进行数据分析,可以获得教室室内最佳混响时间和信噪比的最佳取值,证明了通过构建虚拟声环境来研究教室语言清晰度是可行的.

Abstract:

Speech articulation is the key point of research on classroom acoustics,and speech articulation testing is the most direct approach to decide whether the listening environments are superior or inferior.It is sometimes very difficult to find the required classrooms where factors of listening environment can be regularly changed.Virtual listening environment where reverberation time can be regularly changed can be fabricated by applying exponent attenuation to the signal of actural classroom,so does signal to noise ratio by intermingling noise with the dead signal obtained in anechoic room.Speech articulation tests were done in those virtual listening environments,and then the optimal reverberation time in mid frequencies and minimal signal to noise ratio can be obtained by analyzing the variance of articulation index scores.Results prove that research on speech articulation in classrooms by fabricating virtual listening environment is applicable.