本代码包括vc下对于wave文件的读取，以及声纹识别中mfcc的计算，包括完整的声音预处理和特征计算过程。-Vc under this code, including the wave of documents to read, as well as Voice Recognition MFCC calculation, including the voice of integrity and characteristics of pre-processing calculations.

　在MP3 编解码器的设计中,前向MDCT 与反向IMDCT 是两个计算最为复杂的部分。 提出了一种新的算法,对MDCT 与IMDCT 的实现进行优化,极大地降低了运算量。该算法将 MDCT 与IMDCT 的计算分成三步实现: 前处理, 核心处理, 后处理。其中,核心处理部分由 MDCT 与IMDCT 共用,大大减少了对硬件电路的需求,降低了芯片成本。-In the MP3 codec designs, prior to the MDCT and IMDCT are two reve

Speex 完整源码，包含了编解码及语音预处理模块AEC、NS等-Global Speex source code including encoding/decoding and pre-processing like AEC, NS and etc.