This is a WIP. I am not sure how long it would take me to finish this between school work, thesis and sleeping 🙂 But, hopefully it will be done soon.
In case you dont know what a child leash rig is, it is exactly like a dog leash. One end connects to a hook. Other end is held in the hand.
I am trying to make a real cartoony rig for this. It would be nice to see the following things in it:
1) good standard deforming ability
2) when stretched beyond limits, the leash should compensate by a change in volume.
3) when loose, it should dangle like a rope connected on 2 ends. Will be nice to have an automated swing to the rope. This automated swing should be playable at any position of the belt.
4) will be good if a sine wave can travel through the length of the belt. The shape of the sine wave should be adjustable/damp-able.
For now, this should be a solid start. I will add more features when I think of more.
———————————————————————————————————————————————
June 9 2012
I have spent the last 2 days trying to work this rig. And I am not even talking about the fancy automated movements , etc. Just making the belt deform believably… Phew! its complicated. But I have 2-3 techniques that I think will work. I will need to test these more before I choose a desired method. Also, as I am rigging this for a short, it really depends on the shapes/functionality they need for their short. Will update again once I have more to say and spill here.
For now, the approaches I have are:
1) a set of RP and SC solver IKs that drive a primary joint chain, that in turn drive a secondary joint chain which deforms the belt. (simplest of all, but not great on getting expressive shapes)
2) ribbon rig. Not very complicated. Seems to be getting a lot more deformable shapes. I might settle with this one.
3) SplineIK. A pretty decent rig actually. Just few unexpected behaviours and I have no clue how to fix them. If these are fixed, most definitely going with this approach.
4) Combination of SplineIK and Dynamic Hair Driven joints. Best results in terms of deformations and easy of animation. But am currently waging a war with the settings in the dynamics. Also, this might be over complicated approach to a simpler solution. But, it works so well!!! The rig is a success.. if only the dynamics would work and make some sense….
——————————————————————————————————————————————————————————————————————-
June 14 2012
What I ended up doing…
I drew out a sort of expression sheet for the leash/belt. Based on the drawings, and the tests I did for the above techniques I realized each technique had a problem of its own.
Here are the expression sheets:
Sheet 1:
sheet 2:
Sheet 3:
Results of each tests:
1) the first method of IK rig is still probably the best. It is simple enough and is easy to get the shapes. But, it cannot get all of the shapes above. The main problem being, the length of the belt is being controlled by only one IK chain. So, its not possible to change the shape too much at all. The poleVector allows for a swing of the belt. But thats pretty much it.
2) The ribbon rig is very useful to know. I am thinking, no matter what rig I end up doing , I will use the ribbon rig as an underlying joint structure to finally control the belt. This is great to make those small tweaks to get the exact shapes.
3) SplineIk… This should be working. But, i am not able to get the stretch working at all. It does crazy stuff and I think I know why it is behaving that way. But, I do not know right now how to fix it.
4) I tried a dynamic chain and it worked pretty well. This was the easiest to get the sag in the belt as the two end smoved closer. But, I personally am not a fan of dynamics. This will be the last resort. Moreover, once the belt is being dynamically driven, there is only so much an animator can have control over the shapes.
I am happy with the first technique. The IK chain for the belt, and 2 more IK chains for the hook and handle. But, its time to go back and see how to add on to what I already have and make this rig better.
More Later.
Leave a Reply