You know what would be super cool? If contraption presets could be randomly called via markers on the automation lane.
So, let's say you've got six different presets on your drum machine. Rather than calling specific presets via the automation markers, you could have it randomly call any of the existing stored presets.
Extra cool bonus(es): if there could be some extra optional logic to the randomization. As in:
- Randomly cycle through all presets before repeating
- Randomly cycle through X number of presets before repeating anything
- Preset weighting; where some presets might be called slightly more often or less than others
- Constraining the number of potential presets. So even through you may have six presets on your drum machine in the above example, the randomization is only applied to presets 1,2, and 5 (or any combination of the six).
- Logging the randomly called presets to a text file (or something) so that one could back and note particularly interesting combinations
Given the improvisational nature of audiomulch, I think this could help build upon that. I could see a performer starting off with a lot of randomness and then slowly dialing it in to more specific ranges as the performance evolves.
I'd love to see alternative snapshots, through which audiomulch switches. Especially the drum contraption could benefit from it. Like your point four, Memeories, these snapshots form groups. If you choose a group in the flag, one of the groups snapshot is played.
The advantage is that you can have different groups for introduction, chorus and so on.
The second is that you can group your ideas, and avoid long snapshot-titles.
And the third that you can choose only one particular snapshot, otherwise, and can copy between groups.
But the fourth is you could have variations in timeline snapshots, which is still a problem. (idea:timeline snapshots of le_kzgl, in this forum).
At last: to have something like DNA printed and sliced and played, for the order of random things is very good, for the arpeggiator, too, but I think to complex now at audiomulchs point of development