Got it, Yves .... Let's see what Chris comes up with for you 🙂

Hi again, Yes, the Neo-Riemann comment was mostly just a thought dump of an alternative approach to coding chord progressions which would be very efficient in terms of coding and memory (whilst being fairly unique as functionality). I think I understand your request better, now. You are after functionality that allows a simple chord definition to have vairants which would humanise the chord voicing every time it is played ... kind of like a jazz pianist playing a C major 7th chord one time around a verse, but playing a C major 9th the next go around to add variety, yes? Obviously, Chris knows best here, but on the midicake itself, I would try to create something like this: * Select a chord you like, that you want the ARP to humanize * Setup a chord on track 1, and modulate the start note to change the chord inversion * Setup an arp on track 2 an octave below 1, using the same midi channel. set its time to be the same as track one so notes are played at the same time. Choose a pattern to provide an interesting selection of root notes for your chord from the notes in your chosen chord * Setup an arp in track 3 an octave above track 1. I'd probably choose a more rapid time for this track, but choose a rhythm and bounce to provide sparse clusters of notes (some of which coincide with your chord being voiced) This mights create something with the flavour you're after .... if I've understood you correctly :)

Hi Chris & Yves, Count me as an enthusuatic Midicake Arp user. My primary use case utilises the follow mode, playing a midicontroller to provide the notes/chord for the arp engines to follow. That said, I've also used a sequencer to composer chord loops for the arp that I can use as a dependable vamp during live performance. I've also played with the new chord chains functionality, too, and that'll definitely get used when I pull my next set together 🙂 Ok, so why am I posting? I've been looking into various compositional theories to be able to spice up the sort of music I can perform solo backed by all this wonderful technology .... I come from a background firmly in rock bands 🙂 One thing I've been playing with is Neo-Riemann transforms. If this is not something you've come across, a bit of googling might help, but the general ideas is simply this: Chords are described as combinations of tones on a 'tone net' arranged in a triangular grid. A lot of film music uses this appraoch to create emotional chord progressions that are not diatonic in the traditional sense. The novel thing is that chords are 'transformed' by moving from one postion to an adjacent positon on the grid. There are a limited number of transform types, and they work on any triad (and extensions to a degree). The operation a transform performs on notes is the same no matter what the starting chord is, and the resulting chord has the same relationship to the starting chord, no matter what that starting chord is. So, what I'm thinking is that you could save a series of transforms using far less memory than storing complete chords, and composing using a series of transformations from a starting chord is a fun and new way to create interesting results away from the typical diatonic/ functional harmony approach. Anyway .... I thought I'd share 🙂