in 88, even with peak length and sustain length at 0 (well everything at 0 except attack and peak level), the value doesn’t go back immediately to 0 after attack but stays for one or two samples at peak level.
Just tested it in 77 which has the expected behaviour.
Great. Recently I’ve been using non time-sliced triggers as a quick way to generate animation curves and I noticed a bunch of little quirks, like ease in only or ease out only not doing the full 0-1 range but 0.02-0.98 or something, and changing the units for values at 0 changed the curve (slightly different result if delay is 0 seconds vs 0 frames).
That was in an older build so I’ll test again more thoroughly in the new one.