Feature Analysis of J-core Music Based on Statistics of Python

Feature analysis is a crucial part for recognition of the music type. In this study, the rhythmic, harmonic, timbral and structural characteristics of Japanese hardcore (J-Core) music were systematically quantitatively analyzed based on the Music Information Retrieval (MIR) technology using Python and librosa library. The SoundCloud API was used to retrieve the tags Jcore, J-core and J core, and 30 randomly selected. J-Core songs were analyzed. Thirty tracks in .wav format were randomly selected; then the sampling rate was standardized to 44.1 kHz and merged into mono to eliminate confounding variables. Key features extracted included: mean beat tempo (BPM), most common chords, mean spectral center of mass, and MFCC 1-13 coefficients. The statistical results show that the average BPM of the selected samples is about 179 11 BPM; the spectral center of mass is located at about 4.8 kHz, reflecting the dominance of bright, high-frequency components; 60% of the tracks are dominated by minor chords, which present strong emotional depth; and the MFCC analysis reveals the enhancement of the low and high frequencies and the relative weakening of the midrange frequencies. The MFCC analysis reveals the enhancement of low and high frequencies and the relative weakening of middle frequencies. These results prove that J-Core's unique style is characterized by high speed, bright tone, and minor-key harmonies. This study provides a quantitative framework for the academic investigation of the J-Core niche genre and lays the foundation for the development of automatic classification and recommendation systems based on audio features.