Maybe it depends on the song as well? Classical music is supposed to have a much greater frequency spectrum - meaning, the effects of MP3 encoding becomes apparent when played on high fidelity equipment.
It has greater dynamic range (think volume, this is the domain of the bit rate properly called bit depth) than most compressed studio-produced music but the frequency range or spectrum (think pitch, which is limited on the high end by the sampling rate/2 via the Nyquist-Shannon theorem) is not different if not also clipped off by mixing.