In ultimate vocal remover GUI, you can easily use an integrated method to separate different audio tracks (using a combination of different models in parallel to process the input audio, and then selecting different integrated algorithms to get the final output audio); I found that there are musical instrument sounds (like dongxiao and erxian, etc.) in the vocals separated by a single model. This may be because their frequency spectrum partially overlaps with the vocals, resulting in the separated vocals being unclean.
In ultimate vocal remover GUI, you can easily use an integrated method to separate different audio tracks (using a combination of different models in parallel to process the input audio, and then selecting different integrated algorithms to get the final output audio); I found that there are musical instrument sounds (like dongxiao and erxian, etc.) in the vocals separated by a single model. This may be because their frequency spectrum partially overlaps with the vocals, resulting in the separated vocals being unclean.