This study aims to develop a vocal type identification system for male and female voices based on pitch using the probabilistic YIN (pYIN) method implemented in the Python Librosa library. Pitch, or fundamental frequency, serves as a key indicator in classifying vocal types such as tenor, baritone, and bass (male), as well as soprano, mezzo-soprano, and alto (female). The system is designed to read audio files in .wav format, extract pitch in segments, and calculate statistical features such as mean, median, and frequency range to determine vocal classification. Visualizations in the form of time-based curves and pitch histograms provide insights not only into vocal range but also into the dynamics and expressive techniques used by singers. The results indicate that pitch distribution analysis is an effective and informative approach for vocal classification, with potential applications in training, music analysis, and biometric systems.
Penelitian ini bertujuan untuk mengembangkan sistem identifikasi jenis suara vokal pria dan wanita berdasarkan pitch menggunakan metode probabilistic YIN (pYIN) dalam pustaka Python Librosa. Pitch atau frekuensi dasar suara merupakan indikator utama dalam mengklasifikasikan jenis vokal seperti tenor, bariton, bass (pria), serta soprano, mezzo-soprano, dan alto (wanita). Sistem dirancang untuk membaca file audio berformat .wav, mengekstrak pitch secara segmental, lalu menghitung statistik seperti rata-rata, median, dan rentang frekuensi guna menentukan klasifikasi jenis suara. Visualisasi dalam bentuk kurva waktu dan histogram pitch memberikan wawasan tidak hanya tentang jangkauan vokal, tetapi juga dinamika dan teknik ekspresif yang digunakan penyanyi. Hasil ini menunjukkan bahwa analisis distribusi pitch merupakan pendekatan efektif dan informatif dalam klasifikasi suara vokal untuk keperluan pelatihan, analisis musik, dan aplikasi biometrik.
You may also start an advanced similarity search for this article.