Wish I could say my process was as interesting to read as yours (these are great notes, thanks for sharing).
I guess my track was a mix of accidents and heavy editing. It was built from two separate phrases I recorded, liked, but never fully developed. The first part, with the piano, was run through Arbhar and contains the aforementioned ‘accidents’ in that I inadvertently triggered the onboard mic several times, adding quite a bit of background noise. I got kinda a ‘playing in traffic’ vibe from this error and liked it.
In the second part of the track, the initial layer was something previously recorded, also run through Arbhar with 1 v/o from Orca’s Heart (Ansible) - if I recall, I was “playing” the modulation matrix. Sometime around the end of the fourth minute, I reduced the length parameter to put Arbhar into wavetable mode.