“…Surprisingly, it is also the most limited in terms of the number of aspects of music that it can represent (Fujishima, 1999), since it is designed to be read by hardware, and therefore consists of hardware protocols and instructions, rather than a more "musical" representation. Deeper exploration of different MRLs, their thorough analyses and reviews can be found in (McLane, 1996;Repetto & Selfridge-Field, 1997;Selfridge-Field, 1994). Since symbolic representations already contain many low-level features explicitly, inferring high-level features is less complex, bypassing signal processing altogether.…”