Character Rig

*Check out the model I used for this rig! This model was made by paultosca on TurboSquid.com

https://www.turbosquid.com/3d-models/x-free/864961

• 

• 

• 

• 

• 

• 

• 

• 

• 

The Problem :

For this final project I wanted to create an animation that could be used in a video game. As I started the process of gathering my assets and preparing them for animation I soon came to realize why rigging and animating are divided in the pipeline. I decided I wanted to focus on making a quality rig that would allow me more versatility in animating my character without sacrificing efficiency or readability. One of the biggest problems I ran into while making the rig was keeping track of the many parented elements and making sure I had the correct order of parent to child. When it came time to test my rig with controllers I ended up finding many, many, bugs that needed fixing, many of which the root of the problem was parenting. The second biggest problem I ran into was in the weight painting stage. The character model I chose has many intricate clothing pieces that needed to painted carefully. I had to find a good balance of weight between joints so that when the clothes are pulled they move in a more organic and believable way instead of looking rubbery or being pulled so much the mesh rips.

The Solution :

My first step in the rigging process was to create a flexible rig. One solution to this was to create an IK rig, an FK rig, and a third rig that would have the ability to switch between the two. When done correctly this is an incredibly powerful tool, especially for animators (which I later had the good fortune of finding out myself). The issue with this being keeping up with three rigs and how they are working in tandem. I soon became lost in the ocean of joints and parent constraints, and as expected small details fell through the cracks. As I started the process of connecting the rigs to controllers is when the bugs in my rigs started to make themselves apparent. One of the biggest issues was for some reason on one of my rigs the joints themselves became unparented from each other, essentially rendering my entire rig useless. Once the issue was found this was an easy fix. I found one of the best solutions to this issue was honestly stepping away from the rig for a day and coming back to it with fresh eyes. This allowed me to relook at my hierarchy and connections and figure out what needed to be where and why. Returning to my rig gave me a much clearer understanding of the process, unfortunately this meant deleting parts of my rig and starting them from scratch, but within the afternoon I had almost my entire rig debugged.

Of course any good rig is only as good as the way it’s attached to a mesh, and weight painting was a whole new project in and of itself. One thing in particular with the model I chose was that in some areas where the clothes overlap, the meshes overlap as well. This became a huge problem when weight painting as I would paint the mesh on top, but not be able to paint the mesh underneath. This creates a very annoying problem was when you are moving your rig to test your weights, you get a few random vertices from the other side of the mesh who just randomly decide to join the party. In retrospect, if I was to do this process again, I would split the mesh into individual pieces and skin them separately. This would allow me greater control over each individual piece without worrying about interior vertices. One tool I wish I had discovered sooner that makes weight painting much less frustrating is the Hammer Vertice tool which essential takes any rogue vertex and normalizes it back into place. This tool is a lifesaver when you have that one vertex that is pulled by a joint on the complete other side and instead of trying to figure out which of the many joints it’s attached to, you can just select the vertice itself and fix it. The weight painting overall proved to be a tricky process as I had to figure out how to weight the harder pieces of clothing so that they were affected less and looked more natural when moving. Overall I’m pretty satisfied with the result.

After the hours of rigging and re-rigging, I was finally able to treat myself to testing my rig with a few short animations. Creating the rig itself gave me a much better understanding of the controls and what they were capable of. During the animating process, though short, I was glad I chose to spend more time creating a good quality rig that I can continue to use and animate down the road as opposed to settling for something simpler and more unpredictable.

https://drive.google.com/drive/folders/15Ez_suMuFKzkT-Io7k64qHRMpStdYIsd?usp=sharing