Wavelet Decomposition - Frequency-Scale mapping

I have just started learning about wavelet decompositions, and I have a question that is probably pretty simple. I am trying to decompose an ERP signal using a Morlet wavelet function. I can do the transform just fine, but I need to figure out what scales to use that correspond with the frequency ranges I want to look at. I have tried using the scal2frq() function, but there is one thing that I can't quite figure out about it. Is the sampling period parameter it asks for the sampling period of the signal or is it something else entirely? I feel really dumb asking such a simple question, but without frequency values, I really can't interpret the decomposition in the way I want. Thanks in advance.

Answer 1

[Wayne King]

Hi Zac, It is the sampling frequency of the signal. Also have a look at cwtft(), that is a new Fourier-transform based algorithm for the CWT.

If you look here:

http://www.mathworks.com/help/toolbox/wavelet/ug/bs2gc05-1.html#bs64h74

there is an example of creating an approximate frequency vector for the Morlet wavelet. (Approximate Scale-Frequency Conversions).

Wayne

[Zac]

Wow, that was fast. I will definitely check out the cwtft() function as well. Thanks again.