Repository logo

Investigation of techniques for automatic polyphonic music transcription using wavelets.

Thumbnail Image



Journal Title

Journal ISSN

Volume Title



It has been said (although sadly I have no source) that music is one of the most useful yet useless phenomena known to mankind. Useless in that it has, apparently, no tangible or immediately practical function in our lives, but extremely useful in that it is a truly universal language between human beings, which transcends boundaries and allows us to express ourselves and experience emotions in rather profound ways. For the majority of us, music exists to be listened to, appreciated, admired (sometimes reviled) but generally as some sort of stimulus for our auditory senses. Some of us feel the need to produce music, perhaps simply for our own creative enjoyment, or maybe because we crave the power it lends us to be able to inspire feelings in others. For those of us who love to know “the reason why” or “how things work” and wish to discover the secrets of music, arguably the greatest of all the arts, there can surely be no doubt that a fascinating world of mathematics, harmony and beauty awaits us. Perhaps the reason why music is able to convey such strong emotions in us is because we are (for whatever strange evolutionary reason or purpose) designed to be innately pattern pursuing, sequence searching and harmony hungry creatures. Music, as we shall discover in this research, is chock-a-block full of the most incredible patterns, which are just waiting to be deciphered.


Thesis (M.Sc) - University of KwaZulu-Natal, Pietermaritzburg, 2009.


Music theory.