Obtaining just vocals from a track

Django x2

Django x2

Django x2

Soldato
Joined
28 Sep 2008
Posts
14,222
Location
Britain
Hi all,

I've been messing around with mixes / remixes of songs for years but something I'm really looking to get into is vocal remixes.

Are these vocal tracks obtained under license or is there very good software out there which can split track so you're left with just the vocals (or as near as dammit).

Hoping someone might know
 
There aree plugins for Winamp that can strip the vocals and leave the music. When I tried it maybe a decade ago it came out sounding a bit mushy but it sort of worked. I'm sure you could do the same in reverse, with more or less success depending on how well it was written.

The theory goes that vocals are recorded equally on both the left and right stereo channels, while music is split across the two. If you look for stuff that appears in both channels you should find the vocals.

I think for the proper stuff you usually get a license to use the vocal track in isolation. I remember reading about some rap artists have made their vocals available in the past.

I mean, there must be a way to do it when you have things like The Beastles or The Grey Album, both of which are on very shaky ground copyright wise, so likely weren't made with any official assistance.
 
Last edited:
there's plenty of acapella websites out there. as for plugins for winamp, never used them but they'll no doubt be absolute garbage.

there's also the often used trick of getting the instrumental for the track you want the vocals from and "subtracting" the instrumental from the original track, leaving you with an acapella. the results are usually poor though (any frequencies shared by the vocals and instruments are removed).

i've seen someone on youtube using NI Kontakt to remove the instrumental, the results were usually OK.

your best bet though is a site like acapellas4u (or similar). failing that, you could always try acapella covers on youtube/soundcloud.
 
Thanks guys.

Helpful stuff there. I'll tell you what got me thinking about it.

This:

There is nothing of the original track left behind and I'm not sure the Trap Nation project would be bothered about taking large amounts of effort or money to remove the underlying track and leaving the vocals to be manipulated as much as they have.

Pretty good mix too :D

Also, speeding up vocals without adjusting the pitch. Although that's not the case here, the pitch is up a notch, but I heard a mix of Closer on BBCR1 the other day and it started proper chilled and then dropped but all the time the vocals pitch remained but the speed increased. That was a good mix too.
 
Audacity could do this, including the speed adjustment (without affecting pitch).
 
Keep in mind though, locking the pitch and increasing the speed can work ok (it just removes various milliseconds evenly) but the opposite does not, as stuttering is introduced as additional sound is required to fill the gaps (kinda like frame interpolation in TVs).
 
krooton said:
Keep in mind though, locking the pitch and increasing the speed can work ok (it just removes various milliseconds evenly) but the opposite does not, as stuttering is introduced as additional sound is required to fill the gaps (kinda like frame interpolation in TVs).
Click to expand...
resampling audio is nothing like changing the frame rate on a tv. you've got, at most, 60 frames per second on a TV picture to play with. with audio? usually 44100 frames.
 
Django x2 said:
There is nothing of the original track left behind and I'm not sure the Trap Nation project would be bothered about taking large amounts of effort or money to remove the underlying track and leaving the vocals to be manipulated as much as they have.

Pretty good mix too :D
Click to expand...

That's because there's a very good quality acapella available. Nothing to do with them sitting down at their DAW removing the instruments!

If you'd read my previous posts you'd know where to find the acapella;)
 
slipperydave said:
resampling audio is nothing like changing the frame rate on a tv. you've got, at most, 60 frames per second on a TV picture to play with. with audio? usually 44100 frames.
Click to expand...

Doesn't matter, the principle I was referring to is the same.

Frame interpolation adds fake frames to smooth things out.

Turning a 1 second audio sample into a 2 second sample without changing pitch means 1 second of sound needs to be faked, so it is duplicated a tiny amount every few frames to stretch the sound, causing the stutter. At least that is what Sound Forge does when you stretch whilst preserving pitch/key.
 
Django x2 said:
Thanks, I did see. Signed up :)
Click to expand...
I'd highly, highly recommend that you use Ableton (or Bitwig) for doing these mashups (that's what i'm presuming you're actually doing). I don't know any other program that lets you manipulate audio like those 2. Warp markers are an absolute must for this sort of thing.
 
You must log in or register to reply here.
Back
Top Bottom