Try this version of your simplified patch
As a problem solving (/identifying!) experiment:
-
I’ve added a Length pipe with the
length
set just shorter than your 16hz noise rate (1/16 = 0.0625 =c62ms) -
all notes should go through Scale Remap
-
now try decreasing your LFO
depth
(more duplicate notes)
and increasing the Lengthlength
(note more likely to duplicate before counterpart gets a Note_off) -
watch your Scale Remap get “clogged up” until you click
Transpose C to Base
again to clear it
This illustrates Giedrius design of the pipe.
So your patch will require a bit of a redesign to stop this
PS, you might want to try making your own Sample&Hold instead of using noise…
here’s some discussion and different approaches