Hello, I am trying to determine the frequency of a signal. Samples of the signal are taken from 2 channels at a time. After some filtering, I perform the autocorrelation of each channel, and the cross-correlation of the 2 channels. The autocorrelations, as expected, yield almost identical results, from which we can easily work out the frequency of the measured signal. However, the cross-correlation gives a frequency that slightly differs from that given by the autocorrelation. I am inclined to give preference to the cross-correlation results, since, as far as my understanding goes, correlating independent channels takes advantage of the information contained in each of them, and thus yields better results in terms of low noise and accuracy. Which one should I believe? Autocorrelation or cross-correlation? Thanks for your suggestions, Tao

# Autocorrelation versus Cross-correlation

Started by ●March 14, 2007

Reply by ●March 14, 20072007-03-14

On Mar 14, 6:29 am, "Tao" <taorom...@hotmail.com> wrote:> Hello, > > I am trying to determine the frequency of a signal. Samples of the signal > are taken from 2 channels at a time. After some filtering, I perform the > autocorrelation of each channel, and the cross-correlation of the 2 > channels. > > The autocorrelations, as expected, yield almost identical results, from > which we can easily work out the frequency of the measured signal. > However, the cross-correlation gives a frequency that slightly differs > from that given by the autocorrelation. > > I am inclined to give preference to the cross-correlation results, since, > as far as my understanding goes, correlating independent channels takes > advantage of the information contained in each of them, and thus yields > better results in terms of low noise and accuracy. > > Which one should I believe? Autocorrelation or cross-correlation? > Thanks for your suggestions, > > TaoAre the two sample clocks the same? John

Reply by ●March 14, 20072007-03-14

Reply by ●March 15, 20072007-03-15