Voxengo - Sound with a Soul
Professional VST PlugIns and
AU PlugIns

I've read the manual but still can't figure out exactly what the "Side Mix" knob is doing.  Is it a L/R mixer?  For example, is it doing this:

Out = SideMix * (L+R) + (1-SideMix) * (L-R)

If not, could you describe in equation form what it's doing?  Thanks.

It works like this (SideMix is in the range 0 to 1):

GainS = 1.0 - cos( SideMix * M_PI )
GainM = ( 2.0 - GainS ) * ( 0.5 + SideMix )

It was calibrated this way, because side signal is usually quieter, so this formula adds around 6 dB of gain when SideMix is at 100%.

Thanks.  I assume GainS and GainM are the relative proportions of the (L-R) and the (L+R) signals, respectively, correct?

If SideMix = 0, then GainS=0 and GainM=1, which means Out=1*(L+R).  This would be equivalent to the mono mix.

If SideMix = 1, then GainS=2 and GainM=0, which means Out=2*(L-R).  This must be the 6dB you were talking about.

So, if SideMix = 0.5, then GainS=1 and GainM=1, which means Out=1*(L+R)+1*(L-R)=2*L.  Is the output then normalized?  Or does this get the 6dB boost too?

If SideMix = 0.33, then GainS=0.5 and GainM=1.25, which means Out=1.25*(L+R) + 0.5*(L-R) =~ 1.75*L+0.75*R.  This seems like a pretty hot signal.  I've got to be missing a normalization factor somewhere.

GainM and GainS are direct signal multipliers, they are applied to mid and side signals, respectively.  Of course, the decoding stage does have a "normalization" multiplier.  I'm only talking about how mid and side channels' loudness is transformed.  The M/S encoding and decoding steps are standard.