Hello,
I’ve noticed a consistent issue with high-frequency artifacts when using this repository
When I process the same files with the GUI version of UVR using the same model 2_HP-UVR.pth, I do not encounter these artifacts.
However, when I run separation with the CLI provided by this repo, the output always contains noticeable high-frequency artifacts.
I’ve already tried adjusting various settings (window_size, agg, high_end_process ...) but the issue persists in every case.
It seems like the CLI pipeline here is producing a different result than the official GUI app, even under the same conditions.
Could you please check if there’s a difference in how audio is handled between the CLI and GUI versions?
Thanks in advance!
Hello,
I’ve noticed a consistent issue with high-frequency artifacts when using this repository
When I process the same files with the GUI version of UVR using the same model 2_HP-UVR.pth, I do not encounter these artifacts.
However, when I run separation with the CLI provided by this repo, the output always contains noticeable high-frequency artifacts.
I’ve already tried adjusting various settings (window_size, agg, high_end_process ...) but the issue persists in every case.
It seems like the CLI pipeline here is producing a different result than the official GUI app, even under the same conditions.
Could you please check if there’s a difference in how audio is handled between the CLI and GUI versions?
Thanks in advance!