AD Senior: Warm and mature voice RVC model

Please note that version V3 is Cn_hubert version, so strictly follow the https://klrvc.com/zh/tutorial/3714 tutorial to deploy the model.

Update: 20251201

This update mainly focuses on optimizing Cn_Hubert self-supervision features and improving the 48k sampling rate. You need to strictly follow the instructions above to perform the replacement operation.

v3 is a new version that requires replacing the Hubert file. Follow the instructions above.

The old version was En_Hubert, and the two are not interchangeable.

——————————————————————— ——–

The original dataset of this model comes from the uploader AD senior

This dataset processed a 20-minute dry voice dataset + fusion singing dataset, meaning we found an author with a similar tone to the AD senior and asked him to provide a model that fused dry vocals from several singing parts.

About mature voices: If you need to sing, choose low, gentle songs. Other songs don't work well, but chatting is great.

This is the first time Miaoyin has fused two different datasets into a single model. After testing multiple songs, the results are excellent. Although the mature voice doesn't have a very wide range, its tone at this stage is still excellent.

No reasoning has been

made <!--[if lt IE 9]>document.createElement('audio');<![endif]-->

After reasoning

Singing performance

parameters

training log