Ekstraksi Fitur MFCC pada Lagu Gundhul Pacul

Roger Julian Sitorus
I Gede Santi Astawa

Jurnal Nasional Teknologi Informasi dan Aplikasinya
Vol. 1 No. 4 (2023): JNATIA Vol. 1, No. 4, Agustus 2023

MFCC is an effective method in audio feature extraction, including in song and music analysis. This method involves converting the frequency spectrum of the audio signal into the Mel scale, which is more in line with human auditory perception, and then calculating the cepstral coefficient. The results of the MFCC feature extraction on the song "Gundhul Pacul" show the pattern of the song's spectral and rhythmic characteristics. By using the MFCC representation, it is possible to see changes in energy and frequency patterns in audio signals at various time intervals. The results of the MFCC show that there are 13 resulting cepstral coefficients. However, the number of cepstral coefficients can be adjusted depending on the application and specific needs.