@renegog @crunchydrums @eblomquist
Thanks for getting back to me! Broadly speaking: I have tried all of this over the past year (after discovering my love of the mix output). The issue, invariably, is that phasing matter, which is very difficult to describe in text and which I’m not sure I’ve adequately conveyed here. To give it a slightly more focused go (although my mathspeak is a little rusty…):
I’m looking for some f(x) such that the same x value will not always return the same output, call it f(x, y) BUT also such that the y value as deployed within the function does not radically alter the “character” of the function relative to x taken in isolation. . . that is, ANDing N5 + N6 results in whatever modulation is performed by the result overtaking the subtle modulation of identity through N4 being low in the mix output (you will hear, primarily, whatever AND is modulating, rather than those more subtle MIX movements). Combining the mix output with other signals (through a VCA, logic, or as a modulator) tends to diminish its very special, undulating character (shape + curve ~2 o’clock, although I do tweak these).
This hurtle is of a similar form to using the modulation outputs individually without FM or modulation of intone (since they will always be in phase when used this way, and so give clean divisions of Identity, rather than non-repeating pseudo-random patterns).
Trying to duplicate the ~character~ without the results being strictly deterministic with respect to the ~values~, and without the ~character~ being overcome by the character of some external modulator (such as Maths, modulation from TT, etc…).
This may not actually be a mathematically feasible thing, but I can’t shake the feeling that I’m somehow missing something. that would permit me a larger number of similarly “layered” or “unfolding” modulation sources.
Sorry for being unclear – maybe to illustrate the idea um… Patch mix to a timbrel parameter and identity to a VCA. See how now you only hear Identity modulation of the source (since the sound source is silenced the rest of the time). Similarly, any use of the JF outputs will overlap deterministically (although in more subtle ways). Think about a drone. You want all possible combinations of values for 0 < x < 1 and 0 < y < 1. If using the JF mix output and any control source derived from JFs other outputs, you will not get every possible combination of values because there will always be points at which that value X and Y will always alight on the same value at the same time (or be shifting in the same direction at related rates, cyclically).
Ok… that’s like four different ways of phrasing the impasse that may help inform a solution (or just confuse the hell out of everyone…).
