VOCAL EXTRACTION METHODS

Hoodie wrote:

if your asking about the extra boy plugin lol i not sure i got a windwos ver of it lol they mite have a mac ver not sure here’s the link if u want to check it out lol http://www.acoustica.com/plugins/vst-directx.htm the pro is more money but i got a demo and it worked better lol but the normal one works good to lol and your welcome lol

hey do u have the extra boy plugin?? if so can you send me the file??? pretty please?? lol send it through http://www.yousendit.com

phasing method works fine enough… great post TiMG. Plus just equalizing a vocal out works just as good for me. Some of you guys are missing the point, you only need just enough quality to be able to mix in with another instrumental. In essence your masking the faults with the new mix. If it doesnt sound right, then its because its not in the right key and you probably shouldnt have mixed them together in the first place.

Oh btw maths dude, thats subtraction not inverse. inverse of (x+1)

is 1/(x+1)^-1 but u got the idea anyway.

Cheers!

http://www.djsargex.com

This sound like an idea to try. Is there any software on the market thats affordable that will give me a better quality acappella? Id like to start a few mixes with the pella’ so if the original mix is comin through strong it might not work.

cool, thanks

i tried the methods that were posted and the one that worked best for me was inverting the sound…i use audacity but i was wondering which program would be best for making acapellas. i plan to buy ableton live but want to know if it is as good as it seems.

I dont know if someone pointed it, but it has a way to get your accapela even cleaner.
As we know the phase cancelation is the best way to get a pela out of it’s bed.
But even if this solution is the best, it has still some flaws. (ie: those low level sounds)
They are mostly acceptable, but in certain case, we just want to get rid of them.

First we have to understand where are those sounds/noise/artifacts are comming from. While mastering both the song with the voices,
and the instrumental, the engineer used the same settings that he got on the version with the voice on the version without.
Which means at the verry final stage he used the same maximizer on both version.

Those small sounds comming back after phase cancelation are coming from the version without the voice.
Fortunately they are the result of the same type of maximisation/limitation,
which means if we know how to go back (reverse engineering), we can fix that thing.
the instrumental version has less "things" to maximize which leaves more room for the sounds (artifacts) normaly "under the voice" to shine after phase cancelation.

Now that we know the problem, what is the solution?

The solution, as easy as it sounds, can be tricky.

reduce of 3 to 4 db both version, and add to both a compressor with a reverse ratio, such as 0.5:1, or 0.7:1… (that will "uncompress" them).
Keep the thresholds high enough. If you listen to the "canceled" version, the small sounds are not sounding like snare or hats playing at a low level,
but more like "the peak levels" of those sounds choped of their lower level parts.(they’re mostly sounding like 8bit sounds)
Play on the attack and release settings till you get rid of the small sounds comming back.
When done add a maximizer back to the track, and put it’s level up to what level you substracted at the begining.

nfect

[quote quote="Anoosheh":1jejga85]Hi, I’m new here… I did a quick search but couldn’t find anything direct on the forums, so…

Not necessarily relevant, but I didn’t want to start a new thread, but after vocal extraction, is there any way to convert an mp3 file to midi that anyone knows of?[/quote:1jejga85]

Try using Melodyne I’m pretty sure you can trigger midi in there. I’ve only just got it so still learning the program

thnx

used the phase method, and it works and gives quite a good effect if u got a good instrumental and a high qual song

wow thank you
it sounds difficult

I’m using ableton live, and i can’t seem to figure out the procedure for phase cancellation. You guys said it’s done by flipping the left and right bits of audio right? is this on the instrumental or the main track?

Can anybody send me the invers vst for virtual dj, or uplod it?

Sonify wrote:

I did the phase cancellation in sound forge… All i could get was the kick drum frequency to totally eradicate… i dont know how more accurate you can get.. shanoz.

The instrumental you use has to be a *literal* instrumental of the original – basically it has to be an instrumental created in the studio by doing nothing more than ‘muting’ the vocals. But sometimes an instrumental is also a remix – the levels of other instruments are slightly altered. In this case, the phase canellation method won’t work! And it sounds like you’re in that situation. However, the kick drums have remained the same, so only these have disapeared!

With that in mind, after doing what you’ve done above, you could then use the Knockout plug-in to isolate the centre of the stereo field (losing stuff panned left/right). Knockout also has a low-pass filter which I find is quite good at getting rid of basslines. Not so good at totally eradicating beats (without degrading the vocal as well), but you’ve already done that!

You still won’t get a perfect Acapella (unless all your instruments are hard panned left and right), but more than useable in a mix probably.

Learn something everyday

Great post! Thanks for the information.
It’s much appreciated and will be put to good use.