View Full Version : Using Dynamics vs Keyframing

09-20-2009, 06:03 AM
Hi everyone,

I'm doing this project where I need to animate a creature with a tail (sort of vein stem like a snake) and a head. He's in a hamster wheel and falls forward, causing the wheel to spin about 90 degrees. The creature gets up and makes a clumsy effort at making the wheel spin once again. I'm in between using gravity and rigid bodies to affect both the movement of the wheel and the creature falling down.

I did some test and every attempt has come to the conclusion that the creature is tough to control while gravity is in affect. I still have not addressed the tail, but it simply needs to be elongated as the creature stands, flat while laying down, and then it needs to sort of push the creature back up.

I can send a file if you need to see what I'm talking about
So there, just sort of picture a snack in a hamster wheel. Hope that helps :)

thank you,


09-20-2009, 11:24 AM
sounds like a nightmare to do with dynamics...why not just keyframe it?

09-23-2009, 09:02 PM
Ok that first post was really confusing. Maybe this will clear up what I'm going for. I can't post pictures or share files because it will violate my company's rules and regs.

Well my employer is sort of pushing the use of dynamics in this project.

My approach has changed slightly. Instead I'm going to use a curve and hairsystem to drive the joint chain of the eyeball+vein. With gravity applied to the creature, it should fall unto the hamster wheel (active rigid body with a hinge constraint) and cause the wheel to rock a bit. I'm just unsure about how I'll be able to manipulate the eyeball after it falls down. After it falls, it needs to regain footing, pushing itself up, and fall down again.

Any tips?

09-24-2009, 02:42 AM
Use dynamics to make "the rough cut." Then, tweak it by hand.

Your task is very complex. Although you probably could figure out a way to simulate it perfectly ("push the big red button and voila!") ... it probably isn't worth the time and effort. It's a fantasy that, if you actually manage to do it, won't pay back whatever it cost to do it.

Let the computer produce the simulated movements that it can. Capture each of them separately, throwing away whatever didn't work. Use these as your initial starting motion curves.

Then ... the human touch. Tweak the curves. Add key points as-needed. Let the simulations give you what they easily can, then let your brains do what they can.

09-25-2009, 12:42 AM
Hair doesn't seem to be the best choice to drive the joint chain unless I can get it to not be so darn flexible and string-like.Tried increasing the stiffness but no dice. I may need to go back to before I assigned it to the IK to fix that. I'm controlling the rig using sticky hair constraints so I can get that motion of the creature falling down and coming back up.

As far as modifying the motion curve, I'm going to give that a shot. This 'IS' a tedious method so if anyone knows of an easier one, I'll try it out.

Another issue has sprouted up the hamster wheel moves away from the hinge constraint ever so slightly.

09-28-2009, 11:58 PM
So I've won half the battle so far. For some reason, the eyeball clips with the wheel, but only when it is point constrained to the tail's ikspline. Anyone know why this is happening? The eyeball was became a rigid body automatically when assigned gravity and they did without a doubt collide with each other.
Is there an alternative to the point constrain that might allow collision between the eye and the wheel?



CGTalk Moderation
09-28-2009, 11:58 PM
This thread has been automatically closed as it remained inactive for 12 months. If you wish to continue the discussion, please create a new thread in the appropriate forum.