Real Time Guitar Sound Analyser

OData support
Dr. Fiala Péter
Department of Networked Systems and Services

There are two main subjects of this essay. Firstly to introduce and analyze two of the standard algorithms to pitch detection (ACF, AMDF, HPS, discrete cepstrum): the ACF (autocorrelation function) and the HPS (harmonic product spectrum). Secondly to analyze one special method, the ETR (early transient recogniton), than implement it on FPGA card. Every analyzed method was realized in MATLAB, and tested offline. I analyzed these methods to find out, which is the most effective to pitch detection. The ETR is radically different in this sight, because it needs only one period of the signal to detect its frequency. While the ACF and the HPS has lot of realization in use, the ETR has just one, which was made by András Szalay to AXON. To devise another algorithm for the pitch detection, which based on real time time-range processing, than implement my own ETR algorithm on FPGA, was my final purpose. I investigated the parameters and the possible difficulties of the realization of the methods. I analyzed specially their speed. My ETR algorithm isn’t perfect so far, so I want to go on with this project in the future.


Please sign in to download the files of this thesis.