Discrete Wavelet Transform and frequency bands

I am using DWT to denoise a signal that I have (MER signal), I am importing a data set of type double(1x535840) as a signal into Wavelet 1-D on the Wavalet Toolbox and experimenting with the levels, vanishing moments, and mother wavelets that would give me the best denosing for my signal type. I am finding it hard to understand the frequency bands of the details at each decomposed level. I know the idea of how half the nyquist frequency is divided by two at each level for app(LPF) and detail(HPS) coefficients but since no information about the sampling frequency of the data set was given when loading the signal, I am not sure how frequency band are done in the toolbox.

I read about psudo frequency using scal2freq but it is still not clear to me.

Answer 1

[Louis Edelman]

I had a similar issue and a little obscure googling later, I found the answer. You need to generate the DWT filterbank and then you can query all the time-frequency parameters at each scale.

https://www.mathworks.com/help/wavelet/ref/dwtfilterbank.html

[Shada Sharif]

Thank you Louis, this seems to help with understanding the frequency bands of the filter banks based on the levels of decomposition. I appreciate it.