CAJ | 학술논문

　　系统以单片机和FPGA为控制核心，实现了语音存储与回放系统。能够采集模拟语音信号以及耳机立体声信号，以ADPCM（自适应差分编码）的方式提高了存储器的利用率，语音存储时间可达2 min；基于短时傅里叶变换原理，实现了语音信号的频谱分析与实时显示。同时，利用立体声音频功放播放语音，每声道音量可调并具有静噪功能。此外，系统还采用预加重、去加重、抗混叠滤波等措施，有效地提高了信噪比。语音回放质量良好，存储时间较长。
　　계통이단편궤화FPGA위공제핵심，실현료어음존저여회방계통。능구채집모의어음신호이급이궤입체성신호，이ADPCM（자괄응차분편마）적방식제고료존저기적이용솔，어음존저시간가체2 min；기우단시부리협변환원리，실현료어음신호적빈보분석여실시현시。동시，이용입체성음빈공방파방어음，매성도음량가조병구유정조공능。차외，계통환채용예가중、거가중、항혼첩려파등조시，유효지제고료신조비。어음회방질량량호，존저시간교장。
With singlechip and FPGA as the cybernetics core,the system realizes voice storage and reply system. It can collect and simulate voice signals and stereo signals from earphone and lift utilization rate of memory by the use of ADPCM, which means the voice can be stored for more than 2 minutes. Based on the short?time Fourier transform principle,it can also achieve spectral analysis of voice signals and real?time display. Through using the stereo audio amplifier,each sound track can be adjusted and muted. Furthermore,some measures as pre?emphasis,de?emphasis and anti?aliasing filtering are used in this system to increase SNR efficiently and get good quality of the recorded voice for a longer time.