Hi all, I stumbled across this technique when trying to do mid-side processing
and after thinking about it for a while I think I might have found a simpler way to isolate side that doesn’t end up doubling the signal when recombined. I’m not an expert at this stuff so please let me know if I’m doing something weird - but at the very least I think the output sounds good.
Basically I split the signal with two sends into a mid channel and a side channel, with the effects:
mid: Stereo Expander [Mono]
side: Gainer [Invert R] => Stereo Expander [Mono] => Gainer [Invert R]
The idea being that mono results in (L + R)/2, so the mono projection of a track with R inverted is (L - R) / 2, which is side (well, half of side). If we invert R again, we end up with [(L - R) / 2, (R - L) / 2], which when mixed normally with the mono signal [(L + R) / 2, (L + R) / 2] gives back [L, R].
I usually put custom processing after the second inversion, but if you really wanted to do processing on the mono “side” signal you could put effects before the final gainer (after the mono projection). I also usually send the whole song (by way of an “everything” group that contains every track) through these so I can control more or less the whole song’s mid and side parts.
Edit: an example: 8091 mid-side.xrns