I thought I should give Proteus a go, so … I did. And it does something that is hard for the control freak in me to handle.
It has 3 basic strategies, determined by Lambda. It can produce endless novelty, some randomly attenuated novelty (in other word is repeats for a duration determined by a number of repetitions and a coin flip) and a recording that repeats without change.
I do play around with adding variations to sequencers, but there are limits to what works for the ear, and doesn’t seem aimless. I will use a sequencer like Nysthi Squonk which repeats predictably (and you can save presets with 8Face), and then add a random offset to it’s CV output. This allows you to introduce variations but retain the basic shape of a sequence that repeats perfectly.
Once I start adding random offsets to the output of Proteus it’s just mayhem. Almost better to just let it run and then edit it afterwards to get some repeating blocks to add structure.
When I build it with make dist it’s only 471k. if you look in github the SVG is quite small. I don’t know where they library version are installed, but if they are really 14M something is quite wrong.
I don’t know what @officialdjglitch is talking about artistic panels being huge. If you look on the screen it’s a nice looking panel made of a black background with two circles in it. If you look in the SVG that’s what it is. The SVG does have some inner shadows and gaussian blurs that I think would be ignored by VCV - in any case it looks the same on screen and in Inkscape and Illustrator. Those shadows and blurs must be very subtle.
One of the reasons that I think Proteus is really good, is that it allows you to work with the full spectrum from completely locked and endlessly repeating, to generating something new all the time. And on that whole spectrum what it comes up with and allows for is pretty good to my ears. But yes, for deliberate composition the right thing to do with any generative sequencer, such as also Turing Machine or Marbles and the like, is probably to use it as an inspiration generator, and then capture moments from their output that you like, record that CV and then use it statically from that point on, being able to deliberately manipulate that further. We have some modules that are good at capturing CV.
Certainly a trade-off that comes up all the time. I think I prefer the way most DAWs work, where the audio files remain external, and don’t need to be duplicated every time you do a “save-as”. Is something in VCV smart enough to avoid re-zipping these giant files each time you do an incremental save?
Anyway, pretty obvious trade-off with external vs. internal.
