A study on compare pitch detection algorithms

Cansu Kadi, Seda Gokhuseyin, Yard.Doc.Umut Arioz · 2015

In this paper, three algorithms were choosen by reviewing the literature about pitch detection and implemented on the Matlab environment. Runtimes and error rates of these algorithms were compared with each other. These algorithms are Harmonic Product Spectrum (HPS), Cepstral Pitch Tespit (CPD) and Autocorrelation Method. the observed results showed that the fastest algorithm is CPD and the slowest algorithm is HPS. The least error was achieved at 44.1 kHz sample rate for all algorithms. HPS algorithm become more successful with %99 success ratio in noiseless environment.

Read the paper · More papers on PaperTik