I know that they're equivalent ways of describing a stationary stochastic process and that they contain the same information expressed in different ways, but what kind of datasets would it be preferable to use the spectral density function?

Thanks.