I’m analyzing seismic records and would like to know whether it is methodologically appropriate to evaluate the amplitude of the spectral ratio as a function of earthquake magnitude. Is this approach valid from a spectral analysis perspective, or are there important limitations I should be aware of?