🚀 go-pugleaf

Hi.

I'm trying to work out how to implement both stereo enhancers and 
vocal erasers (in a possibly mistaken view that they are related).

I'm trying to extract "pure" left and right signals from a stereo signal.
Assuming the siginal I've got has something panned hard right, something
panned hard right, and something in the centre, I can write the sum of
the left and right channels as:

	raw(left)+raw(right) = true(left) + true(centre) + true(right)

when summed to give a mono signal.

Now, if I calculate: raw(left) - raw(right)

I get: 		     true(left) - true(right)

Where the centre signal is absent, and the left and right signals are
present relatively out of phase (this may be useful as a vocal eraser of
poor quality).

But I have been unable to use combinations of sums and differences of 
raw(left) and raw(right) to isolate the pure left and right signals. To
describe what I mean by "pure" I mean that something panned exactly 
centre would be absent, but something panned 10% to the right would be 
present only in the "pure right" signal (i.e. not at all in the pure
left signal) and would be attenuated proportional to its centrality
(i.e. dead centre = 0% original volume, halfway between centre and right
= (e.g.) 50% original volume, hard right 100% original volume).

Ummmm, I hope I've made myself clear here. Is there any simple way of
doing this?

Thanks in anticipation,

Ross-c