Form and Space

creating CG 3D since 2001

  • Connect @
  • Linked in
  • Vimeo
  • Twitter

Adaptive Rigging

Matthew | April 16, 2010

1 Comment

After a conversation with my student, Jake, I was reminded of some old research and experiments I did between April and July 2007 with adaptive rigging.  I spent my summer off from teaching trying to recreate Bernard Haux’s excellent rigging work.  I don’t think my work was nearly as clean as Mr. Haux’s, but I was tapping into some good ideas for character freedom.  The traditional animators of Disney were free to draw and not be limited by the technology’s or the technical director’s limitations.  They could make marks on paper, over and over again.  Their two primary tools are the pencil and the paper.  Simple.  Of course, I’m oversimplifying the traditional animation methods, but it is clear they were not bogged down by lists of tools or the failed design of a rig.  One of the most challenging principles to include in a rigid 3d character is appeal; I’ll refer to appeal as the quality of the contour lines.  Another challenging principle to include is squash and stretch; both automatic and manual S & S controls.  In fact, most 3d I see floating around the web lacks this essential principle.  In fact, often the animator is blamed for the lack of principles.  Yet, it’s not necessarily the animator’s fault, and is most likely the rigger’s engineering of the character’s movement failed.  It took me years to realize I was leaving out squash, even though I had achieved stretch.  Sigh.

This rig design would adapt to the needs of the shot and scene.  Allowing an animator to stay true to the storyboard and character designs.  This gives tools for the animator to “free” the character from conventional methods of animating a rig.

In this concept demo I had a laundry list of options I was trying to achieve.  Again, inspired by Bernard Haux’s demos.  The arm had to have:  1. IK/FK switching for the whole arm,  2. Independent IK/FK blending for upper arm and lower arm,  3. Independent scaling of upper and lower arm,  4. Each node can be “torn away” from it’s parent,  5. Reshape the arm with a curve and objects along the curve,  6. A free elbow achieved by having IK enabled for both upper and lower arm ,  7.  Blend in standard corrective shapes,  8. Independent stretch for upper and lower arm.  9. Finally, Re-sculpt the mesh at any point during animation and key the sculpt.  Take a look.

Adaptive Rig Prototype

The most successful outcome from studying Mr. Haux’s work, which I didn’t originally intend, was my step into object oriented programming or OOP.  After creating my first successful “arm” rig I decided to automate the creation of this complex problem, so I could modify and recreate it without the need to repeat a list of steps in the software.  Unable to automate the entire rig due to time constraints I completed one facet of this problem.  That facet being the rig that runs along the curve.  The “back bone” of my rig.  Curve rigging has led me to some wonderful work, and has even changed the way I model.  Now it’s possible for me to draw any curve or curves, and apply a rig to the curve with the click of a button.  I build several features into this rig, and more recently have added sine wave auto-animation feature into it.  Probably the strongest feature is the built in squash and stretch based on the length of the curve, and based on the compression of the object along the curve.  Take a look at the curve auto-rig.

Rig From Curve

Programming the remaining rig setup was the next step, but I lost steam and time.  Why did I lose energy?  Well, I attended SIGGRAPH 2007 shortly after completing the demo reel.  While there I learned from a few Dreamworks and Lucas Arts riggers that Python is THE programming language of choice in the 3d entertainment industries.  That’s when I lost steam.  I was doing ALL of my programming in J-script.  Pfft.  (I discovered years later, after trying to find info on how to use OOP in J-script that it simply is impossible to have true OOP in J-script).   I stopped pushing this rig after running out of energy, and running out of time, and a new semester was about to begin.  Regardless, here’s my demo reel from July 2007.

Demo Reel 2007

My dream is to return to this problem, and complete it in both XSI and Maya.

My most recent use of the curve rig is in Red, the Ball with a Tail.  You’ll find superior squash and stretch with both the ball and the tail.

One response to “Adaptive Rigging”

Leave a Reply