Forums

Something like cross correlation for the time domain

Started by Unknown November 2, 2013
 From time to time I have to do cross correlations of signals with 
different sampling rates. The problem is that I do not know the ratio of 
the sampling rate, e.g. because longer recordings are taken with 
different crystal oscillators for the sampling rate.

The idea is to pass two recordings to an algorithm and get the pitch and 
the delay as result. Strictly speaking the result is two dimensional, 
giving the correlation for each pair of pitch and delay. But I am mainly 
interested in the maximum.

I have no idea how this is called. Any ideas where to seek for that?


Marcel
Cross Ambiguity Function is what you are looking for. 
Similar things are used for time and frequency estimation in many communication systems. 
On Saturday, November 2, 2013 6:52:45 AM UTC-4, Marcel M�ller wrote:
> From time to time I have to do cross correlations of signals with > > different sampling rates. The problem is that I do not know the ratio of > > the sampling rate, e.g. because longer recordings are taken with > > different crystal oscillators for the sampling rate. > > > > The idea is to pass two recordings to an algorithm and get the pitch and > > the delay as result. Strictly speaking the result is two dimensional, > > giving the correlation for each pair of pitch and delay. But I am mainly > > interested in the maximum. > > > > I have no idea how this is called. Any ideas where to seek for that? > > > > > > Marcel
On 02.11.13 20.54, asaidhiraj@gmail.com wrote:
> Cross Ambiguity Function is what you are looking for.
Indeed, this was the right search term. Thanks. Marcel