My Journey into Animation

Revisiting a Rig

by on Dec.29, 2013, under Uncategorized

Earlier this year I did a rig for a mechanical cat that was posted on CGtalk in their Rigging Challenge section. I think I have mentioned it before but it has been in my rigging example reel for a while. As I have uncovered more about rigging, I realized that things should have been done differently. In the last version, I used a series of IKs in groups which I programmed using set driven keys. In some cases it worked fine but in other cases it created errors. Some I had fixed by controlling settings that are probably best left alone.

A few weeks ago, I decided to revisit it but use what I had learned since my last attempt. Starting with the spine, I wanted to try a flexiplane which proved difficult since the spine isn’t completely straight. Still, I came up with a way of doing it by using a curve from a spline IK with blendshapes to get it to adjust to the correct shape. I created the flexiplane without any issues but the issues came in getting the joints to do what I wanted them to do since each piece only moves in specific ways. Some can only rotate the X, the Y, or the Z axis. In the end, I settled on just a spline IK with the tail section following the main body.

Next up were the legs. My setup would suggest an IKFK system but the way I programmed the lower section, FK wouldn’t really work that well. Still, I tried a few different ways of setting up the joints to move how I wanted and hit some snags. I couldn’t get the leg to move in all directions by locking some rotational axises. As with the spine, the parts can only rotate along one axis. My solution was to lock the foot controls X translation and use a Z rotation control at the top of the leg instead which works just fine but is one more step that I want out of the system. The legs did require a correction system since the position changes when the shocks at the top are moved. A simple 2 joint chain where the two rotation points were and an IK tied to the lower section worked. I just used the rotation generated from that to correct both the top and bottom Z rotations.

For the ankle/foot section, I used locators instead of joints. I then put those locators in a series of groups that had pivots at different locators to create the correct movement. Some actions required moving higher locators which caused issues as lower ones needed to remain where they were so they were split up and given separate groups with the same pivot points and linked together. Combine this with multiply nodes that countered the initial rotation made the movements possible. The back leg even has an extending ankle that allowed me to discover a few new controls I didn’t know about but let me adjust the pivot point based on the ankle length. Testing that did lead me to realizing that some parts weren’t staying connected because the split groups weren’t changing with the leg. For example, the Y rotation for the foot wasn’t changing position/angle when the ankle section was bent so it would detach.

I think that is enough talking. I screen captured my demo of the rig and put it in my rigging reel below at the end. I am in the process of doing an animated presentation of it now.


Leave a Reply

Looking for something?

Use the form below to search the site:

Still not finding what you're looking for? Drop a comment on a post or contact us so we can take care of it!

Blogroll

A few highly recommended websites...