Algorithm info: Submitted by: Bas Curtiz
Version: UVR v5.5.0/Patch_12_22_22_23_44/Win
Advanced MDX-Net:
Chunks: Auto
Chunk Margin: 44100
Volume Compensation: 1.035
Denoise Output: On
Spectral Inversion: Off
Advanced Demucs:
Segments: Default
Shifts: 1
Overlap: 0.25
Chunks: Auto
Chunk Margin: 44100
Enable Chunks: Off
Split Mode: Off
Combine Stems: Off
Spectral Inversion: Off
Based on the fact Kim's vocal fine tuned model scored highest in non-ensembled.
However, this is an older model (November), than she uploaded at 2022-12-04.
I asked her, but she doesn't have the newer model anymore, since she's still training it.
Time Elapsed: 01:03:06
Date added: 2022-12-25 |
