Comments on the spectra of Pisot numbers