Feature Extraction and Musical Personality Analysis of Piano Artistic Performances

• Published in: Applied mathematics and nonlinear sciences Vol. 9; no. 1
• Main Author: Liu, Shuai
• Format: Journal Article
• Language: English
• Published: Sciendo 01.01.2024
• Subjects: 68T05; Feature extraction; Piano string vibration equation; Short-time Fourier transform; Timbre reconstruction
• ISSN: 2444-8656; 2444-8656
• DOI: 10.2478/amns-2024-1283

Starting from the basic process of piano music recognition, this paper focuses on the feature extraction process of piano art performance. In the feature extraction of piano art performance, the fundamental frequency characteristics of piano music signals were analyzed by using short-time Fourier transform theory, focusing on the feature extraction of musical pitch, time value, and tonality, and the spectrum analysis of piano music signals and the reconstruction of piano timbre were completed. The piano performance work, “Autumn Moon on the Peaceful Lake,” is used as an example to analyze the musical characteristics and personality. The overall structure of “Autumn Moon on the Peaceful Lake” is a typical three-part structure: the key is the D palace seven-tone mode, the key is single, and the marking range of performance strength is ppp-pp-p-mf-mf-f-ff. In the three versions played by Yin Chengzong, Lang Lang, and Chen Jie, the average tempo of the performance was 31.5 beats, 26.5 beats, and 26.8 beats, respectively. The performance process of different pianists is a “slow-fast-slow” layout, and other passages are added to the individual free processing to further enhance the charm of the work.