Datasets:
Update README.md
Browse files
README.md
CHANGED
|
@@ -46,6 +46,8 @@ The recordings emphasize harmonic isolation, spectral clarity, and controlled ov
|
|
| 46 |
Developed by **Harmonic Frontier Audio**, this preview follows *The Proteus Standard™* for dataset provenance, transparency, and ethical AI use.
|
| 47 |
Learn more about the Proteus Standard → https://harmonicfrontieraudio.com/proteus-standard
|
| 48 |
|
|
|
|
|
|
|
| 49 |
It offers clean, consistent audio files capturing overtone singing across stable fundamentals, moving harmonic trajectories, vowel-modulated resonances, and controlled expressive gestures.
|
| 50 |
|
| 51 |
If you find this dataset useful, please consider giving it a 🤍 on Hugging Face to help others discover it.
|
|
|
|
| 46 |
Developed by **Harmonic Frontier Audio**, this preview follows *The Proteus Standard™* for dataset provenance, transparency, and ethical AI use.
|
| 47 |
Learn more about the Proteus Standard → https://harmonicfrontieraudio.com/proteus-standard
|
| 48 |
|
| 49 |
+
Full dataset details and licensing information are available at: https://harmonicfrontieraudio.com/datasets/overtone-singing
|
| 50 |
+
|
| 51 |
It offers clean, consistent audio files capturing overtone singing across stable fundamentals, moving harmonic trajectories, vowel-modulated resonances, and controlled expressive gestures.
|
| 52 |
|
| 53 |
If you find this dataset useful, please consider giving it a 🤍 on Hugging Face to help others discover it.
|