Extracting Acappellas - Ableton

franklyshocked

I've seen lots of little posts on this recently on various boards and they were all more about removing the vocal to get a clean music track (Probably for Karaoke or some other rubbish) rather than extracting the acappella's like most people seem to want.

Its probably not that different in most DAW's but Ableton is my weapon on choice so here you go:

Set up 3 empty audio tracks in Session view
Rename track 3 as submix.
In the in/out section at the bottom of tracks 1 and 2 route both into submix instead of to master.
On the monitor section of Submix select "in" rather than auto.
Take your original full track and copy it side by side into tracks 1 and 2.
Go to your audio effects. Drop a utility device on track 2.
On the utility working from bottom to top, set it as follows:

Click phz l and phz r so that they are both on,
Set width to 0.0%
leave Panorama as c
in the drop down box select SWAP

If the 2 copies of the song are side by side you can launch both at the same time with they will play simultaniously into the submix channel.

In most audio production the vocal is hard panned to the centre channel.
With the phase of one track being swapped you will have a cancelling effect of the left and right parts of the track leaving a relatively clear vocal.
Enable record on the Submix track to record your vocal.

No its not perfect, but if you can't get the acappella anywhere else then at least you can get a half decent one to make a start with.

Hope that helps.

Any of you Logic or Cubase lads want to throw up how you do this process on your platforms?

0neo8

Please reveal how its done on Logic Express

chordtype

Cheers for this. Gonna try it now.

krd

The whole trick works by phase cancellation - that is the Sine wave of the bit of vocal you'd like removed added to an inverted version of the wave.

It's easier to get rid of a vocal then to cleanly separate the vocal. Vocals are not always absolutely on centre - and usually they have stereo reverbs which makes it harder. Sometimes you can actually here it when a sample has been used with the vocals removed - you'll hear a shadowy reverb of the original vocal in the background.

If I were you I'd practice with something like Goldwave. You may never get the vocal perfectly clean - but the background noisy will be in tune and and rhythmically in time to the vocal.

If you search hard enough you may be able to find the stems for what you want on line. But if you listen to say vocal samples used by the prodigy, you'll notice they're really dirty - they haven't been dirtied up, that's as clean as they could be got.