thanks for the review. as for the missing noise-source: one of my own most wanted "feature-request" is to allow for arbitrary samples in the osc-section, not just single-cycles. so we may as well load a noise-sample then. at some stage, we'll hopefully see that.
hmm - free-running oscillators? really? do you want that? i must say, that i am personally very concerned about exact reproducibility of the output. i don't really like when every note has a different pseudo-random attack transient. i mean, technically, it would be easy enough to implement but i'm sceptical about the desirabilty.