low energy transistions are the ones which are the most liable to be useful for understanding the behaviour of your material in "normal conditions"
in a terribly rough first approximation, the behaviour of your material is dominated by its first energy gap (HOMO - LUMO for instance) and therefore the interest in the low energy end of the spectrum
but other transistions are interesting too ...
it will depend of what you expect from your spectroscopy ; for instance, if you want to identify a material, you might need a full fledged spectroscopic signature !