Dubstepforum.com

Right.

I know these "removal" programs don't really leave good results..I've tried most of em, So i aint even bothering with tryna find a decent one.

I understand u cant remove the vocal, in effect just turning down the background music without compromising the vocal TOO MUCH.

I was wondering if anyone knew any good EQ techniques i could possibly use to do this instead. (Theres alot of EQ Wizards in the production forum), Or an actual program which delivers a "decent" enough result, so i could go with the EQ thing AFTER.

If it was as easy as getting the ACC's.. trust me i would.

There are alot of Coldplay/Fray (Indie genre?) bits i wanna fart around with. But this genre is flipping hard to find a resource for.

Help !

Cheers in advance lads

D.

I've never done it myself but as i understand it the method used sometimes is running the instrumental at the same time with an inverted phase so it cancels out the instrumental element leaving just the vocal, but the results aren't perfect obv

if you want to remove the vocal, the best thing to do is either get a cheapo karaoke machine, or a used alesis vocal zapper.

otherwise (and i'm waaaaay to hungover to explain the detail) split the track into a mid-side matrix, and flip the phase on the mid channel-- you'll lose most of the lead vocal... and a good amount of the kick/snare/bass as well.

if you want to isolate the vocal, hi and lo pass the track and limit the piss out of it. OR, try the trick mentioned above, if you have the instrumental.

good luck-- trying to remove a part of a song is like trying to un-make soup.

TeReKeTe wrote:
good luck-- trying to remove a part of a song is like trying to un-make soup.

hahaha. totally.

1. get a stereo track

2. split it in to two mono tracks

3. invert one of the tracks

4. done

It dose not always give good results but it is the easiest way

KEBNOA wrote:1. get a stereo track

2. split it in to two mono tracks

3. invert one of the tracks

4. done

It dose not always give good results but it is the easiest way

I really don't understand how that would work!

Surely that technique would just remove all of the common frequencies between the two stereo channels and leave the different ones?


As far as I can see, if you don't have accapellas to invertify, its just an EQing MISSION. Dropping out frequencies etc.

its my understanding that this sort of trickery will only work on really older tracks (think early beatles) where there was true stereo separation of the vocal tracks??

I mean I guess heavy eq'ing might work but you're squashing loads of stuff and vocals aren't this segregated freq range that you can just take out and have everything else that was in the mix still sound fine..

I've heard home-made "instrumentals" and most sound like rubbish bootleg karaoke tunes...


best bet is to snag every vocal-less second from whatever bit you are trying to toy with and re-piece it together in an editing program - surely coldplay (coldplay?) has exposed riffs that play once or twice before the vocals come out on top of it right?

I dont know I'd rather cheat at it than deal with any kind of production mumbo-jumbo that involves the words "matrix" in it - thats just over my head

gravious wrote: I really don't understand how that would work!

Surely that technique would just remove all of the common frequencies between the two stereo channels and leave the different ones?


As far as I can see, if you don't have accapellas to invertify, its just an EQing MISSION. Dropping out frequencies etc.

this works for me with indie music because the vocals are normally the same on both channels and the guitar, bass , drums, etc are normally to the left or right a bit
so you do lose some sound of the guitar, etc and you do get a mono track but I think it works better then EQing the hell out of it.

thanks for all the responses lads, im guna try out a few of these methods.

I know it was a long shot, but im glad it aint totally doubtful!

Cheers

U guys for president ! ! !

unlikely wrote:I've never done it myself but as i understand it the method used sometimes is running the instrumental at the same time with an inverted phase so it cancels out the instrumental element leaving just the vocal, but the results aren't perfect obv

Correct.

1 .Load the original track in an audio editor.
2. Load the intrumental of the same track
3. Zoom in full and line up both tunes, they need to be completely in the same position so use something like the peak of a snare to get them full in line
4. Highlight the instrumental version and 'invert phase'
5. The tunes will cancel each other out due to 180 degree phasing and all that is left is the vocal and any other differences between the tunes.

That sounds bless, now its a case of finding the instrumentals of said genre.

Which is, quite possibly, harder than finding the accapella's.

Guna try re-angle mi search though, wish me luck.

Cheers again!

Vonboyage wrote:That sounds bless, now its a case of finding the instrumentals of said genre.

Which is, quite possibly, harder than finding the accapella's.

Guna try re-angle mi search though, wish me luck.

Cheers again!

If theres a few clean bars with no vocal you could loop it up and try using that on one of the verses, should work fine for heavily quantised/sequenced stuff but you're probably fucked if its guitar music as the timing is often less rigid