David NG McCallum schrieb:
I've noticed that the wet output of freeverb~ sounds like the signal's always panned centre, regardless of how the input's panned. Is this a mistake, or the nature of the reverb model?
It's how the reverb model is implemented here. The comb filters get fed with a mono signal (sum of both channels), only the allpass filters are fed with independend channels.
Olaf
PS: you could change
input = (inL + inR) * x->x_gain;
// Accumulate comb filters in parallel for(i=0; i < numcombs; i++) { outL += comb_processL(x, i, input); outR += comb_processR(x, i, input); }
to
// Accumulate comb filters in parallel for(i=0; i < numcombs; i++) { outL += comb_processL(x, i, inL); outR += comb_processR(x, i, inR); }
to get rid of this