Introduction - If you have any usage issues, please Google them yourself
The code is extracted from the SPEEX news out of the echo part of the source it is taken to operate in the frequency domain, the effect is better " by the test, if the original voice message and a reference for the echo between the voice of the clock is not synchronized easily occurred in the response message algorithm in the phenomenon of anti-ulcer divergence. compressed package file containing the test voice: Input_exsample_stereo.pcm: 8k16bit linear pcm, including dual-channel voice. vleft.pcm: 8k16bit linear pcm, from Input_exsample_stereo.pcm separate the left channel audio. right.pcm: 8k16bit linear pcm, separated from the Input_exsample_stereo.pcm the right-channel audio.