Still very WIP, but I’m pretty proud of myself. This little demo is actually pretty fun to do.
(I probably look ridiculous in real life during this)
Pretty cool idea. I guess you did this in C++?
Yes. It might be doable in blueprint, but I don’t think so. Basically, ragdoll the character mesh, then puppet it around using UPhysicsHandleComponents on the hands, pelvis, and head. The thing is, it takes a bit of Quaternion math to translate the data properly from the Razer Hydras (or whatever motion control device you are using) and the Rift to ensure everything rotates correctly without gimbal lock. Blueprints don’t normally do quaternions though, and I’m not sure if rotators would work.
This is pretty darn cool, I like the swinging!
My curiosity got the better of me a while back and I did something a bit similar, but in my case it was more of a climb-on-any-geometry with kinetics
Not sure what to do with it yet, but its quite interesting how climbing in vr already feels like climbing, except you’re not as heavy!
Use blueprint node combine rotators it uses quaternions internally to avoid gimbal locks.
Ah, that looks pretty similar, minus the ragdolling. Right now I use two traces coming out of the hand and I compare the dot product of the trace hit normals to the friction of the physical material to decide if it is climbable.
Interesting, with friction it sounds like it has a lot of flexibility like being able to handle inclines and such! The hand does some weird things in that video though=p, is it the physics taking priority over hydra orientation when it’s attached?
I went for a simpler approach of only being able to climb on objects that have a clearing above them (using box traces). Mainly thought about emulating a tomb raider style wall climbing, jumping and grabbing ledges, pulling yourself up etc.
These checks are done on the collision geometry so more complicated models will need more accurate collision models than typical games have today.