I would like to separate the bass from this track in the attachment. Then I will either play a new version or modify the existing one. I have tried it in Lalal.ai, luckily here it can unmix the bass completely, and it can do a really good job. However, here and there you can catch the remains of some harmonics, or “holes” where a gap can be suspected in the remaining content, but it is not a serious problem. If you don’t know the bass was unmixed, maybe you would not notice. However I want to do a job as perfect as possible, can you do it in Spectralayers better than in Lalal.ai?
I don’t really know. I’m sure you already know, yet if you do not:
Higher FFT sample sizes bring lower frequencies into sharper focus. The issue then becomes higher FFT sizes increase number crunching time…so, the higher your FFT size setting, the more lag user will wait through to navigate the spectral display.
