Separating vocals and soundtrack for movies/series

[robercano]

Hi UVR community,

I love UVR and I've been using it for vocals separation on movies and series for educational purposes on dubbing movies here in Spain. While using UVR with vocals/instrumental separation has yielded pretty good results, it is not perfect and some sound effects get also extracted into the vocals stem. I'm currently using the following models:

• VR Arc: 1_HP-UVR
• MDX-Net UVR-MDX-NET Inst HQ 3
• Demucs: v4 | htdemucs

I've also searched this forum a bit and haven't found much more.

Is there a combination of models/stem pair that would allow me to have more control or better results for vocals in movies/series?

[Alpen1]

Yea I'd like to know this for anime. I wanted to make some fan edits.