turbid briar
Before Start :
- it's a conference in ENGLISH, so the dataset is indeed in English
- For best results, I recommend you watch the reference video I've put up just after this post, and not recommended to use TTS (as I may have done). Knowing that he's Hungarian, he has quite a pronounced accent when he speaks in English, which even the feature ratio at 1 can't fully replicate.
- Maybe a new version with more epochs ? (to be seen)
Last Update : <t:1703712097:R>
- Model URL :
- https://huggingface.co/rayzox57/AratoAndras_RVC/resolve/main/AratoAndras_v2_250e.zip
- Version :
RVC V2.0
- Pitch Extraction Algorithm :
- RMVPE
- Epochs - Steps :
- 250- 7.5k
- Dataset :
- ~ 00:09:00 (Waking up as a meme-hero - Andras Arato -TEDxKyiv)
- Recommended Usage :
- Speech
- Search Feature Ratio :
- 0.75 ~ 1.0 (Even with 1.0, the model still has difficulty replicating its Hungarian accent, so you may have to force the accent when recording.)
- Pitch :
- Logic Pitch ( 0 = Man / -12 = Women )
- You can adjust if you found an better result
Previews :
Preview_TTS.wav :
Contains External Effects : No
Pitch : 0
Feature Ratio : 1.0
Preview_Cover.wav :
Contains External Effects : Yes
Pitch : -6
Feature Ratio : 1.0
eager canopy