Description
Trained from game assets' samples directly over mono channel. 50ms padding between samples. Only trained 180 epochs as after the regression becomes positive and the model is overtraining. I also tried larger samples but it seems that it's not immediatelly clear that more data is better in fidelity so far. If someone requires to fine tune this further let me know
Comments
No comments yet. Start the conversation!
Add a comment
Samples
New
Classic
1. Singing
Male
English
2. Singing
Female
English
3. Singing (Dry)
Female
English
4. Singing (High)
Female
English
5. Singing 2
Male
English
6. Singing (Dry)
Male
English
7. Singing (Dry, High)
Male
English
Pitch
Users also tried
More to explore
Loading more