Harmonic-Frontier-Audio/Subharmonic_Phonation_Vocal_Fry_Preview

这是一个高保真的人声数据集，展示了次谐波发声——一种复杂的扩展人声技术，用于多种世界歌唱传统——专为AI训练、音乐研究和创意音频开发而设计。数据集包含高质量的、权利清晰的录音，具有详细的元数据，以96 kHz / 24-bit的单声道格式录制。数据集旨在用于AI语音建模、扩展人声合成、表现力音色研究和非线性发声分析。README还详细介绍了数据集的内容、录音说明、使用方法、许可信息和未来发展规划。

A high-fidelity vocal dataset demonstrating subharmonic phonation — a complex extended vocal technique used across multiple world singing traditions — designed for AI training, music research, and creative audio development. The dataset includes high-quality, rights-cleared recordings with detailed metadata, recorded at 96 kHz / 24-bit in mono format. It is intended for AI voice modeling, extended vocal synthesis, expressive timbre research, and non-linear phonation analysis. The README also outlines the contents, recording notes, usage, licensing, and future roadmap.