What are the differences to using a skeletal mesh on mobile compared to pc?
The reason i am asking is that my game works great on PC. The Phat / skeleton simulation works nice. I set it up so that its nice and tight.
But on mobile its almost like the skeleton no longer obeys its constraints, and joints are flying wildly in different directions. As if the constraints are no longer applied the same.
Edit: I now managed to get it to behave the same on pc and mobile.
My confusion is this.
There are 2 nodes. Set all Bodies Physics Blend weight. And Set physics Blend Weight.
What i have experienced is i need to add both in the event graph. Occasionally it would use one and not the other.
If i add Just All bodies Physics blend weight then it will work on 3 launches. Then no longer on the 4th. Acting as if blend weights are set to 0.
So it works intermittently. Thus i added Set physics blend weight.
To me they both seem to be doing the same thing. Except both if used single, works occasionally and fails occasionally but if used in conjunction with one another works fine.
So i guess my question is, what is the difference between the 2 nodes.
Also i would like to find out why is gravity stronger on mobile than it is on Pc.
For example if you have a ragdol, on pc it would fall at 10 feet per second. But on mobile it will fall at 20 feet per second
After more investigation it seems that the values of blend weights are interpreted differently from pc to mobile.
Which is what is causing the illusion that objects are heavier.
Right now i am trying to find a work around for this. So far what i have come up with is to set the gravity lower… and increase the blend weights of the skel meshes.
This is creating the best effect thus far. My only concern with this is, iteration takes long because on the previewer it behaves completely different from on a device.
As such in order to test one change you first have to launch to device to test.
I think we recently found a limitation in the way we were blending when using a value that’s not 0 or 1. I believe this is fixed for 4.6 but I’ve e-mailed the dev who’s been doing that work. She’ll be able to give more info on possible differences, though I wouldn’t expect there to be any intentional differences.
I don’t see there should be any difference. The issue I can think of is low framerate, which can cause unstable simulation, but substepping should have helped on that. Other than, it still should behave same. Does this happen when you make them sleep after simulation?
I suppose it will be very hard to trigger exactly same pose on both platform, but I would test something like this. Have very simple mesh to simulate (just drop from the top) and make sure its simulation result is same in both platform. Now add blending weight for both and check how the result is.
Hi guys n girls.
Thank you. please feel free to close this one.
I never made them sleep unfortunetly. Also the current other issue i have with rag doll now being… Faulty makes this problem something i wish i still experienced