Wavosaur

[thormes]

It would be a good feature if you could you a 'reverse remove vocal' effect. It`s pretty handy when it comes to extracting stuff like movie dialogues and such.
_________________
-= hoy brujerizmo pa ti satanizmo =-

[FordMustang]

i dont understan

[thormes]

Waveosuar can REMOVE vocal from a file by cutting out the mono signal and leaving the rest right? A very useful effect is the opposite, which let`s you keep the vocal and cuts out the rest of the signal. now you get it?
_________________
-= hoy brujerizmo pa ti satanizmo =-

[FordMustang]

but is it possible ? i can't believe it

[thormes]

This function generally works poorly, but for example I often need to extract a dialogue from a movie, which often has background music.
Such an effect removes the stereo field from a wave and keeps the center channel.
Since speech is in 99% of the case a mono signal, placed in the center, it remains untouched while the background music get`s roughly erased. And I couldn`t stress 'roughly' enough. This usually is enough though for me to cover the remaining background with my music and the vocal is heard clearly enough.

:)
_________________
-= hoy brujerizmo pa ti satanizmo =-

[FordMustang]

cool !
i wnat it

[rajarameswar]

[R.A.W.]

What? I don't get what you mean to be honest..
_________________
ThinkPad T500,Core 2 Duo T9600 (2x2,8GHz)/3GB RAM/Radeon HD3650/FireWire 410/WinXP SP3 32bit
AMD Phenom II X4 3.2GHz/4GB RAM/GeForce GTX460/ESI U24 XL/Win7 x64
Ableton Live 8/energyXT2/APC40/MPKmini/LaunchPad/TriggerFinger
http://www.bitmud.com

[rajarameswar]

[R.A.W.]

open your wav in wavosaur and use process >> vocal remover


and raja, sometimes it might help NOT to use abbreviations in every second word to make people understand a post.
_________________
ThinkPad T500,Core 2 Duo T9600 (2x2,8GHz)/3GB RAM/Radeon HD3650/FireWire 410/WinXP SP3 32bit
AMD Phenom II X4 3.2GHz/4GB RAM/GeForce GTX460/ESI U24 XL/Win7 x64
Ableton Live 8/energyXT2/APC40/MPKmini/LaunchPad/TriggerFinger
http://www.bitmud.com

[FRESHA]

Right.... one way to try this.....

1) Open up the track you want to extract the vocal from
2) use the 'vocal remover' tool in Wavosaur
3) invert/flip the sample
4) save as <track-name>-instrumental.wav
5) select all... copy
6) open up the 'full' vocal track again - select all
7) Edit - paste special - paste and mix

This should in tun leave you with the vocal (or most of it)

what this does is phase mixes the two together - the only difference being the vocal in one and not the other. This can be useful, but it isn't great - you may end up getting a 'squeltchy'/'lo-quality-mp3' like sound, because of the similar frequencies overlapping the central vocal and other instruments pnned around - because most tracks are not hard panned this is why the vocal will 'lose' some/most of it's sharpness quality

However... take a look at E-MU's Emulator X3, which actually aims to achieve this technique. They do it with a track by U2 (trying to extract Bono, with pretty good results too) - similar practical approach

[poivre]

Thank you FReShA :

I think it's possible to get good result once tou have the original & the vocal removed one with a spectral substracting instead of invert & pastemix

I use Kn0ck0ut VST or ReaFIR, have to tweak a lot parameters, as FFT size etc

[mg33]