What I meant to ask about crosscorrelation: Suppose I have 2 signals r[n] and y[n] each of length N, and zero pad them to double their lengths and compute Y[k]R[k]* using FFTs in matlab. And then I zero pad that product with 8N zeros BEFORE taking the 10N-IFFT to get the crosscorrelation. It seems my correlation results are worse, but I cant understand exactly why. any help is appreciated! This message was sent using the Comp.DSP web interface on DSPRelated.com
Zero Padding a DFT
Started by ●February 2, 2005
Reply by ●February 2, 20052005-02-02
quick one wrote:> What I meant to ask about crosscorrelation: > > Suppose I have 2 signals r[n] and y[n] each of length N, and zero pad them > to double their lengths and compute Y[k]R[k]* using FFTs in matlab. And > then I zero pad that product with 8N zeros BEFORE taking the 10N-IFFT to > get the crosscorrelation. > > It seems my correlation results are worse, but I cant understand exactly > why. > > any help is appreciated! > > This message was sent using the Comp.DSP web interface on DSPRelated.comThe usual technical verbal shorthand (jargon) hides a lot of details. Zero padding of a time sequence before taking its Fourier transform is fairly straightforward to describe and understand. BUT what is zero padding of a sequence of Fourier coefficients? When done slowly and in full detail the Fourier transformed operation of zero padding transforms into an interpolation operation which takes a moderately complicated explanation. Perhaps you need more that about 1/3 of a jargon filled sentence to say what you intended and what you did.
Reply by ●February 2, 20052005-02-02
Reply by ●February 2, 20052005-02-02
Sorry it was to help understand some results from a lab i did. But I have already since figured it out on my own! Thanks :) This message was sent using the Comp.DSP web interface on DSPRelated.com
Reply by ●February 2, 20052005-02-02
me again wrote:> Sorry it was to help understand some results from a lab i did. But I have > already since figured it out on my own! Thanks :) > > This message was sent using the Comp.DSP web interface on DSPRelated.comPlease don't keep us in suspense; explain a little. Jerry -- Engineering is the art of making what you want from things you can get. �����������������������������������������������������������������������
Reply by ●February 2, 20052005-02-02