快速傅立叶变换-->关于语音对比的第三方软件设计算法
3星 · 超过75%的资源 需积分: 9 177 浏览量
2010-08-12
09:15:20
上传
评论
收藏 79KB DOC 举报 
// -*- C++ -*-
// automatically generated by autodoc
// ========== HEADER FILE src/fft/fft-default.h: ==========
// Default FFT implementations
inline void fft(double *fr, double *fi, ulong ldn, int is);
inline void fft(Complex *f, ulong ldn, int is);
// ========== HEADER FILE src/fft/fft.h: ==========
// (complex) FAST FOURIER TRANSFORM (FFT)
// typical format of arguments: (double *fr, double *fi, ulong ldn, int is)
// fr := pointer to data array (real part),
// fi := pointer to data array (imag part),
// ldn := base 2 log of array length
// is := sign of exponent in fourier kernel
// naming conventions:
// some_fft() := fft implementation with blabla-algorithm
// some_fft0() := fft for zero paddad data
// (i.e. second half of the input data is expected to be zero)
// ----- SRCFILE=src/fft/fhtfft.cc: -----
void fht_fft(double *fr, double *fi, ulong ldn, int is);
// FFT based on FHT
// ldn := base-2 logarithm of the array length
// is := sign of the transform (+1 or -1)
void fht_fft0(double *fr, double *fi, ulong ldn, int is);
// FFT based on FHT
// version for zero padded data:
// fr[k],fi[k] == 0 for k=n/2 ... n-1
// ldn := base-2 logarithm of the array length
// is := sign of the transform (+1 or -1)
void fht_fft_conversion(double *fr, double *fi, ulong ldn, int is); // aux
// preprocessing to use two length-n FHTs
// to compute a length-n complex FFT
// or
// postprocessing to use two length-n FHTs
// to compute a length-n complex FFT
剩余10页未读，继续阅读
资源评论
gxulgc2013-02-27学习FFT的话可以参考，如果是语音对比的话，这个就是骗分的资料
acejoy2012-09-04毫无用处，文档里面一堆杂乱的代码，纯粹骗分的。
