Currently, SpectraLayers can only separate all the vocals from the background music, but it cannot distinguish between the lead vocals and background vocals. The only software that can perform this function is AudioSourceRE’s Demix Pro. Please consider adding this feature to SpectraLayers in the future.
8 Likes
There are other software programs that separate lead vocal from backing vocals like Demix Pro … so I assume the next upgrade of spectralayers Pro will have that feature … otherwise everyone will wonder about steinberg & co …
1 Like
I’m also interested/wondering in if it is possible to Unmix multiple singers.
What happens if there is no “one lead”? (E.g. different persons singing the verses)
What if I have a vocal group of 4 singers? (soprano, alt, tenor, bass, can those be separated)
2 Likes
I generaly grab/extract the entire mixed vocals first, just to get them away from everything else.
If the recording is super-clean, 4pt harmonies are often then easily extracted into separate mono tracks via Melodyne.
For me anyway. I’m used to the look of the blobs.
I’ve never tried the entire approach with Spectralayers.
2 Likes
Great comment … Great request!!! … other software claims they do it and the software costs a fortune so they must be able to separate the voices …
I did it here.
I wasn’t planning on openly discussing this but I actually started working on my own algorithms specifically targeting this kind of content, however because of numerous issues I’ve come across and because I am too proud to ask for help I abandoned that project altogether.
2 Likes
Good comment, Spectralayers cannot separate lead vocals from harmony vocals at the moment and it also cannot separate saxophone solos from a mixed track. Hopefully they will get there one day.
RipX is closer to being able to do both these things, but its not quite there yet. And when Steinberg has perfected this, how about the ability to create Ai lead vocals like Audimee and Kits Ai? - then Steinberg would gain all the Ai customers!
2 Likes
Actually with Ultimate Vocal Remover it can also be done, and it is totally free. Although it would be great to be able to do it in SpectraLayers, as a good professional and paid program (and not especially cheap) that it is.
1 Like
I don’t see any training models in Ultimate Vocal Remover that’ll pull a vocal out of stacked vocals…which is what is being asked for.
If the Ultimate program can indeed pull Judy’s voice off of the stacked four parts below…and isolate each of the four singers on clean separate mono layers… automatically…
I’m all ears…
2 Likes
No, stacked vocals no, but it can distinguish between the lead vocals and background vocals, with the karaoke method. In the previous answer I was referring to the main topic of the post.
Haha. I had Judith Durham in my studio recording three songs with me for the Sydney Olympics back in 2000. She was great to work with. What a unique and wonderful voice. And lady.
1 Like