NVIdia Flex Feedback into Actor?

Hi all,

I hope you are doing well, this is a question on Flex I believe that would be useful to see on the forum for many of us: I am trying to feed brand new positions in my Actor for its particles using the Feedback input. So far the new positions get “forced-in” without a hitch but their collisions are barely working after that.
I find that I can only feed very little increments to the original values to have collisions work well on the particles of my actor.
That leads to a big question: is it even possible to “force-in” brand new position values that can be very different into an actor using the Feedback input. Doing that would allow this:
regularly refreshing the “contents” of the Actor when you have way too much particles for Flex to handle at once. I am doing that refresh loop according to a distance calculation to the solid Actor that is hitting the particles. Hoping somebody has a clue, and that it’s actually possible without restarting Flex.
I am also wondering about using Pre-roll but not quite sure if that is my missing piece.
Oh, and I only need about 1000 particles at one time.
All the best! P

Hi all,

So I’ve been way further on this but not finished yet. Building slowly a “buffer” that feeds particles from a large point cloud into the Flex loop, and out of it when they have been collided. Really tough for me but formative. Happy to take advice if any. Since Flex has a limited budget of “balls” it will be very useful I believe. P