本文阐述了一种实用的语音信号实时延时方法、对信噪比作了估算,描述了采样频率的选取方法,并对延时时间作了推导。它和一般语音处理器的不同之处主要在于时序控制的方式相异。实践证明,通过对逐个存贮单元的先读后写方法,能使语音信号不断输入,延时信号不断输出,从而逼真地模拟了回声,扩大了用途。 This paper presents a practical voice signal real-time delay method, which estimates the signal-to-noise ratio, describes the sampling frequency selection method, and deduces the delay time. The difference between it and the general speech processor mainly lies in the way of timing control is different. Practice has proved that through the memory-by-one read-write method, the voice signal can be continuously input, the delay signal is continuously output, which realistic simulation of the echo, expanding its use.