Automatic Identification of Frequencies of Musical Notes in Polyphonics Wave Files

Carlos Alberto Lopes da Silva
Gilberto Arantes Carrijo
Luciano Vieira Lima
Antônio Claudio P. Veiga

Federal University of Uberlândia


This paper deal with the problems that can be found when someone wishes automatically identify frequencies of musical notes recorded in WAVE format. The same approaches the problem of the analog to digital conversion and of the sample rating of the files in this format when analyzed through Fast Fourier transform. This transform generates a significant amount of lateral lobes around the existent frequencies in these files, taking the professional to identify musical notes not existent in the original files. Like this, this paper presents optimum values of windows to minimize such problem. An appropriate window is not just an efficient and definitive solution to this project. This way a forecast method,refinement and identification is presented here to identify musical notes in files in the WAVE format.