一种基于分析的非语言声音信号合成方法及其在多媒体系统中的应用

音频媒体在多媒体系统中起着非常重要的作用，但由于其存储需要巨大的空间，使得其应用受到了一定的限制。本文提出了一种基于分析的合成方法以便在多媒体系统中对非语言声音信号进行有效的存取。这种方法首先对非语言声音信号进行谐波分析，然后对所得到的时变振幅与频率等声音信号参数进行分段线性化。

AN ANALYSIS-BASED SYNTHESIS OF NONSPEECH AUDIO IN THE MULTIMEDIA SYSTEMS

Audio medium plays an important role in the development of the multimedia systems. However,its huge memory demanding makes the storage somewhat difficult. In this paper, we propose a technique to efficiently access the nonspeech audio under the environment of the multimedia systems. This technique performs harmonic analysis of the nonspeech audio signals and then the resulted attributes including time-varying amplitude and frequency functions are piecewisely linearized. By storing only these highly reduced parameters, we obtain a much compact represenvation of the audio signals. Additive technique is used to resynthesis the sound. We can flexibly control the acoustic characteristics of the sound that are expressed in an explicit way. This property is very useful if we want to use nonspeech audio signals as encoded messages.