I want to make sure I am understanding the filter banks correctly in DWT decomposition.

If my sampling frequency is 30kHz, I would need to input data with 30k samples which later will be decomposed

level 0 - 0-15k Hz

level 1 - 0-7.5k & 7.5k-15kHz

level 2 - 0-3.75k, 3.75k-7.5k & 7.5k-11.25k, 11.25k-15kHz

and so on...

I do think I might have gotten this wrong because say I need to zero power-line interferences (60Hz) but my signals of interest are in the low frequency range then I would need a min of 7 levels to get to 0-117 Hz and that still does not have very high resolution in frequency, and might be hard to implement in real time due to computational power.

Therefore, does the number of samples to decompose related to the sampling frequency or it can be any number below the sampling frequency to obtain higher frequency resolution?