How can Overlap-add technique be done

4 次查看（过去 30 天）
Ijeoma Madu
Ijeoma Madu 2012-12-17
Hi:), I would like to perform the overlap-add technique but it seems to have errors. any ideas
load handel
N=length(y);
L=8192;
b=[0.2];
a=[1 -0.8];
[hn,t]=impz(b,a);
M=length(hn);
Nx=floor((N+M-1-1))/(L)
H = fft(hn,N) %% (zero-padded FFT)
i = 1
while i <= Nx
il = min(i+L-1,Nx)
wt = ifft( fft(y(i:il),N) .* H, N);
k = min(i+N-1,Nx);
w(i:k) = w(i:k) + wt ; %%(add the overlapped output blocks)
i = i+L;
end
  2 个评论
Walter Roberson
Walter Roberson 2012-12-17
What errors do you perceive?
Note: you might want to consider using "for" instead of "while"
Ijeoma Madu
Ijeoma Madu 2012-12-17
编辑：Ijeoma Madu 2012-12-17
Thanks Walter. i get this error and not so sure what this means
Undefined function or method 'w' for input arguments of type 'double'.
Error in ==> asp2 at 20 w(i:k) = w(i:k) + wt ; %%(add the overlapped output blocks)

请先登录，再进行评论。

请先登录，再回答此问题。

采纳的回答

Walter Roberson
Walter Roberson 2012-12-17
Initialize your matrix w, with
w = zeros(1,Nx);
before you start the loop.

更多回答（0 个）

类别

在 Help Center 和 File Exchange 中查找有关 Fast Fourier Transforms 的更多信息

标签

Community Treasure Hunt

Find the treasures in MATLAB Central and discover how the community can help you!

Start Hunting!

Translated by