-
To get acceptable results from phase cancellation you need to have two almost identical quality tracks… one that has the vocals and the other the instrumental. The only track you need to work with at this point is the instrumental, which will be the track you invert.
Please note that inverting a sample is very different to reversing a sample. Inverting a stereo sample flips each channels waveform upside down where as reversing a sample… Well, I’m sure you can figure that one out!
For the sake of clarity the ‘illustrations’ below will be displayed as a single (mono) waveform. The principles for each channel of stereo sample are exactly the same.
[code:11fr8h17]
Waveform before inversion:
/\ /\ /\ /\ /\
/ \ / \ / \ / \ / \
/ \/ \/ \/ \/ \Waveform after inversion:
\ /\ /\ /\ /\ /
\ / \ / \ / \ / \ /
\/ \/ \/ \/ \/[/code:11fr8h17]
As you can see above, peaks in the waveform before inversion become troughs after inversion and troughs before become peaks after.
To avoid confusion it is also worth noting that the inversion process does not flip the waveform as a whole. To do so would indeed invert each channels waveform but it would also swap the left and right channels which would prevent the phase cancellation from working. Inversion treats each channel independently… The left channel stays as the left channel but is flipped upside down, like-wise the right channel stays as the right channel but is also flipped upside down.
So, now I’ve explained a little bit about the terminology and the differences between these terms, so you (hopefully!) understand whats happening, you can go ahead an invert the instrumental only.
Different software may each call the process of inverting a waveform something different but for now we’ll assume that you know ‘what’ the inversion function is called for the software you use. For example in SoundForge it is actually called "Invert/Flip" and is located at Menu -> Process -> Invert/Flip.
So now you should have an inverted version of the instrumental which means its time to combine the inverted instrumental and the original vocal track.
How you do this depends upon the software that you use. In some software you have to mix the two tracks together to hear the results of the phase cancellation and in others you can create layers (which when played together create the same effect as having mixed the two tracks). In practical terms a sample editor that can use layers is more flexible as you can move the alignment of the two tracks backwards or forwards until you achieve the optimal phase cancellation results. Software in which you have to mix the two samples together before you hear the results requires a bit of trial and error but the use of ‘markers’ (if your program has them) can make the aligning process a little less fraught.
Now you should have something that resembles an acapella but you may find that some frequencies haven’t been cancelled out perfectly… for example some elements of the bass may be left behind. If you’re lucky you can EQ those rogue frequencies out without affecting the actual vocal element too much. In other cases if the frequencies that have ‘escaped the net’ are around the same as the vocal it will not be possible to EQ out the rogue frenquencies and you may not be able to use the acapella… unless you find some way to either mask them or incorporate them into whatever you intend to use them for!
Hope that all helps guys!
Shades
AdSense 336x280
The forum ‘D.I.Y Acapellas’ is closed to new topics and replies.
