No takers yet? It’s a very complex algorithm structure. I don’t have any insider information, but I can tell you how I approach Shimmer myself; right or wrong. The exposed parameters are interconnected and interactive. So it’s difficult to discuss one control in isolation, without first explaining some of the other controls that affect it.
Reverb is first in line, and always present. DECAY and SIZE at 0 set the minimum amount of influence (and also determine the ‘shape’ and diffusion of the reverb). It branches off to a DELAY line, that’s always at least 50 mS (for processing); up to 1000 mS. Think of the LOW-DECAY, HIGH-DECAY, and MID-DECAY as a multiband mixer to the inputs of the A-PITCH and B-PITCH shifters. One or more of these frequency bands needs to be above 0 to send some amount of signal to the pitch shifts.
PITCH-DECAY is the ‘master’ control after the pitch shifters & the blend of frequencies sent into them. It’s also the feedback level for the path starting with DELAY, through ‘frequency mixer’, & into the pitch shifters. Then all that is fed back into the reverb up front, within a range from 0-100. So it’s the PITCH-***** level, plus the *****-DECAY level for the whole feedback loop.
I don’t think the PITCH: FREEZE / PITCH+VERB: FRZ variations here have a counterpart in any of the other algorithms. Shimmer’s FREEZE is somewhat like ‘INF’ values in the other algos, but not quite. It won’t hold infinitely; the frozen sound degrades and disintegrates fairly quickly. I suspect because it’s fed back into the reverb, and ‘blurred’, or carved up by running through the frequency bands repeatedly. In addition, input signal still runs through the active reverb.
And PITCH: FREEZE disconnects from the pitch shifters. Any DELAY time fed back around is at the original input pitch. In other words, it can be used as a ‘conventional’ delay line that follows an omnipresent reverb, with regeneration and filters. There’s just no pitch shifting at this particular value.
PITCH+VERB is closer to the ‘FREEZE’ setting found in other algorithms. Input signal comes through ‘dry’, and whatever reverb / pitch / frequency spectrum that’s captured is held indefinitely, recycled back through the reverb, DELAY, etc. You can still modulate some of the parameter controls while it’s in that ‘frozen state’ (using SIZE as an obvious example).
