Poor Audio Quality After Removal

[rorschachhhhhh]

The resulting audio quality from a 3+ minute suno generated file seems to have a lot of artifacts added in various places. It would be nice if there was an option to only do "Complete Metadata Removal" + "Spectral Watermark Detection & Elimination", and skip the "Statistical Pattern Normalization" + "Human-Like Imperfections" steps. Perhaps that would maintain the quality better whilst removing the watermarks?

[geeknik]

I agree, the current output is pretty terrible. I'm in the midst of a refactor. \m/

[Nanasaki741]

Would be more accessible if partial removal is applicable