Automatic Classification of Musical Instrument Sounds
Generally speaking, no real instrument produces an accurate repetition of the same basic pattern. It turns out that these changes in microscopic scale are significant in indicating the nature of an instrument . Both the frequency- and time-domain representations of the typical steady state of an instrument are difficult to obtain, and are moreover not as informative as one might expect. Thus, creating a data base that consists of musical sounds requires a great deal of concern, remembering that both the way of playing and recording method may cause differences in a analyzed sound.
KeywordsFeature Vector Musical Instrument Automatic Classification Recognition Score Discretization Process
Unable to display preview. Download preview PDF.