Fourth Year

Fixing Problems

To say there’s been a few issues with the rig is probably an understatement. I don’t know much about rigging or the technical aspects to character creation but I’m learning on the fly. It usually resolves around a process of me trying hundreds of different things and pushing every conceivable button until something good happens, and then realising there’s a much simpler and faster way to do it that produces better results.

IK FK controller issue

One of the main issues that has been frustrating me for a while is the IK FK switch for the left arm in particular. The others seem to be holding pretty solid so far, which is surprising since it’s my first attempt at this. With the issue only happening in one instance and with a mirrored version of the same thing in the right hand, I just needed to spot the difference. However for a long time I was doing this blind, just trying different things and pushing buttons in the hope that I fixed it. Conveniently Maya has a hypergraph for both connections and hierarchies that I wasn’t aware of, meaning I can follow a road map of each and every connection and see where the path alters.

Here is the version before the fix, as you can see there is some differences between the left and right versions. In the hierarchy hypergraph the point constraint comes in at a different level, which makes it react differently. And in the connection hypergraph the point constraint isn’t found at all. I originally thought this meant I had not point constraint IK Handle to the IK controller but I think I may have instead made the constraint in the opposite order. The most likely reason being the time and energy I had spent into making the rigs made me overlook something by mistake. Going into autopilot for some of the more monotonous tasks means mistakes can be made, especially when I had already completed one half of the character, its just a matter of repeating the exact same process. Mistakes were made.

And here are the graphs after the fix. Identical. This is a mistake that is often avoided through the use of scripts which automatically repeat actions and take away the monotony of some rigging aspects. I however was probably doing this at some point in the early hours of the morning after 2-3 weeks of rigging and testing.


Another big issue I’ve been having with the character rig is the deformation in movement. My thought was that my topology and weight painting was not good enough, whilst this may be true I took a long time to try and fix this. My guess would be that there has been over 30 variations of the topology and models, some with minor variations in edge flow to better mimic anatomical form and replicate its movement, others with larger changes such as blocky proxies and abstract models to better understand the process or to better see the movement.

Here are just a few of the examples mentioned, some changes seem so minor as to not affect anything but the direction of edge flow is vital to the character movement being believable. Since my film involves realistically inspired movement and models, the degree of accuracy and level of polish needed is quite high to sell the movement. Perhaps leaning into the abstract or cartoonish more would have been the sensible decision, the level of quality I am aiming for is difficult to achieve for a one person project and more so given my level of knowledge and experience. But these are the types of projects I want to work on, the level I strife to achieve, and to aim for anything less feels like a concession. In all my work I tend to follow a similar pattern, and it mostly begins with drawing. I learn the rules first and break them after. I first learned anatomy and capturing the likeness in a realistic manner before learning how to abstract reality, to exaggerate and accentuate it. This is a process that has worked in every other aspect of my creative journey, so it makes sense to do the same with animation.

Through the process of learning how to best think about topology and character design, I have learned that I really love that process – of designing a character and designing their movement through form. Studying anatomy and applying that to movement mechanics has been a fascinating part of this project and one I’m sure will continue to be a side project always. In a way the biggest problem I encountered in my project has been the part I have been fascinated by the most, and one I am looking forward to exploring further. The culmination of all the things that make a character what they are, not just in their visual design, but their biomechanics, anatomy and design coming together to inform the manner in which they stand and walk or move, the very essence of their personality. …I can sense some abstract character designs based on animal anatomy coming in the future.

Personal ramblings aside, a lot of the deformation errors were fixed by a different form of skinning in the end, though the topology did help considerably.