Hello everyone!
I have been searching everywhere for a good vocal extraction program, but I haven't been lucky enough to find one. I checked some websites that claim to remove them for you automatically, but none of these seem to work properly, as in most songs I try, it takes some other frequencies along with the vocals I cant cut even with an EQ. I have also tried RX8 for this (thank god for sister site), but it seems to have the same problem... Can anyone recommend something that works as they claim for this, in order to clean cut vocals from songs, or do you have any website to recommend where I can find sharp clean acapellas from songs as if they were recorded in the studio?

 

splitter.ai ?

or try to find acapella online

 

hello jimrantos,

here>www.voclr.it for acapellas......
have fun.

 

Izotope RX8

 

7 free acapella generators : extract vocals from any song
https://afroplug.com/7-amazing-free-acapella-generators-extract-vocals-from-any-song/

 

Last edited by a moderator: Apr 21, 2021

Spleeter is fantastic though can be a bit fiddly to get started with.

I personally find it easier than RX because it has less options whereas RX is a whole audio manipulation toolbox.

 

I compared spleeter and rx8 , rx8 takes longer time to render but its always better imo I use spleeter for fast stuff that I dont care if its the best quality

 

I have used RX8 a lot, but never managed to get a clean vocal, especially on modern music.

 

A.I will take care of it

https://ezstems.com

https://melody.ml

https://phonicmind.com/

 

Acid Pro' s Stem Maker
Xtract Stems

 

this splitting technology (pompously called Artificial Intelligence) is still a bit buggy. I'm afraid that any of those methods use the same algorithm, thats why they sound so similar. Depending on the original material you will have more or less artifacts, There is no way to have a clean acappella stem for now

 

There is no app momentarily on the market that can efficiently isolate vocals or achieve a perfect phase cancellation without background noises.
The handling process to extract vocals from a mix without artifacts or instrumental sound bleeds is still far from our reach.

I hope sooner then later a smart Dev comes up with the perfect tool.
I"ll be the first one to support it and to buy it as soon it get released.

 

Last edited: Apr 22, 2021

As you noted, it is hard to extract vocals without some bleedthrough noise. I did some experiments with the SpectraLayers trial recently and after removing the vocal layer and playing back this supposed 'instrumental' I was surprised to hear the vocals were still coming through somewhat. I have yet to try any of the free stuff on the market and see how it compares to SpectraLayers.

 

As someone noted above, RX8 Rebalance has many options and can work sometimes but I've found Neural Mix Pro by Algoriddim is pretty darn good and super easy to use.

https://www.algoriddim.com/neural-mix-pro

 

website that uses AI (spleeter) to isolate vocals and instrumentals on any song. www.audiostrip.co.uk

It is completely free so worth a try.

 

cool, but how long will it be "completely free"? cuz i remember a time when splitter and melody were completely free.

 

RX8 but if you're using modern music reduce the volume before applying Rebalance (just choose one of the presets, its very straight forward) however you mostly will never get a perfect acapella, the best ones ive got are from old records where drums and music are not so limited. Before using RX8 try to find the acapella as some are leaked or recorded straight from a vinyl release... Another option www.acapellas4u.co.uk (seems down atm). If you have an instrumental mix and normal release both in wav you could try to invert phase too... there are many ways to skin a cat

 

Last edited: Sep 25, 2021

PhonicMind online and so good
but paid )

 

I think a near-perfect separation is just impossible. Ooray, I'm a genius stating the obvious
That being said, Deep Learning (AI) techniques are indeed different and will probably be better than traditional ones in the short/medium term.

 

yes, it's working, but only if instrumental was exported from the same project (with just muted vocal tracks) and all fx/instruments/master rack are the same.

but more and more copyright owners making simple trick to stop it. instrumental without vocal parts sounds in the middle frequencies "poorer". they just use a little of eq for middle frequencies and after this mission "inverting phase to extract vocal" - impossible.

maybe pay to vocalist and record max similar vocal part at studio is the best way? =) at least quality of this vocal part will be more-more better - 100%!