i can make them hit a target just fine, simple enough i wasn't going to do a video on it. now a better idea is a video that shows arrows being deflected at an angle threshold then bouncing off. unfortunately we dont have actual shape collisions of the instances so having them collide with the ground properly wouldn't work.

BTW all of this was inspired by a colleague of mine, brandon riza, an amazing fx artist who has done everything i have shown before and better than i have done. these demonstrations i hope illustrated how easy it is to do what used to be near impossible for the average user in the legacy system. there is nothing here that most other modern systems can't do already :)

look for warhammer : rnd : arrowflocking...
brandonriza.com/3DVisualEffects/HTML/3DVisualEffects.htm

i haven't seen any way to control the offset of the instance from the particle center, but that doesn't mean its impossible.

i already mentioned actual shape collision of the instances isn't in the first release. the arrows would be approximated by spheres and wouldn't look right for arrows.

i was playing around with instance shapes today and noticed that it doesn't seem to use "sphere" primitive for RBD collisions on objects that are long and narrow. it looks like it actually used bounding box like it should. my initial tests proved otherwise and the documentation says its approximated with a sphere too. so i might revisit this with RBD collisions on the ground

Yeah, the manual says it uses approximated sphere but obviously it's bounding box, not bounding sphere. I think what manual says is wrong.

But we need 'actual shape mode' anyway :D


Great videos, I've seem them countless times.
Thank you, Steven.

you can offset the location of the smoke particles, you should be able to get the 'emitlocation' attribute, then the position, modify it, and set it back all on the execute on trigger port. if not you should be able to do it right after wards in the smoke particles graph... but why would you do that? flaming arrows are flaming because their tips are on fire, not the tail end of them.

in order to get the alpha channel from that image as a per particle attribute you will need to do some type of closest location qeury