Emily Ward

Please find my final outcomes below!

The final stage of this project was showcasing the rig using the animations I had created. I decided to use a turntable in Unreal to showcase my animations, since it renders quickly but also would utilise the fact that this rig is functional in a game engine. I used this stylised forest scene as the setting since I wanted all of the focus to be my character and its animations, and created a sequence in unreal to record my shots. Within Unreal itself I used my UV snapshots from Maya to create an emissive layer above the character’s textures to broadcast the wireframe of the model, as well as creating a simple grey texture to showcase the model itself. I had some minor issues with the Unreal sequencer since it performs differently to Maya’s animation but ultimately I was able to use unreal to render some high quality shots of my rig in action.

One minor issue I encountered is that Unreal engine does not allow for renders of bone joints, so as a result that information needed to be directly recored from Unreal’s editor, which means those shots may be lacking somewhat visually. Despite this however, I think these shots provide some good understanding of how the rig is functioning in engine. I think that comparing these different shots together provides a good understanding of my work and progress thoughout this project.

I wanted to demonstrate some of the finer capabilities of the rig, as well as demonstrate some amount of personality. I decided to create this coin magic animation in the style of an overwatch highlight intro, which are usually examples of the individual character’s personality or interests. This animation allowed me to utilise the finger and upper body controls in more detail than the other two animations. I really enjoyed creating this animation, as there was some amount of technical challenge in making the magic trick look plausible, and managing switching parent constraints on the coin as it moved from hand to hand.

As a criticism of my rig, I found that the orientation of which I have rigged the arms in particular may not be exactly ideal- in the rigs I have enjoyed working with the most, the elbow usually needs only to be rotated on one axis of rotation like a hinge joint. However I found that I needed to rotate the elbow somewhat in a non-natural way in order to have an appealing animation which indicated that either my model or my rig could be improved in this area. However ultimately I think the outcome is appealing and fulfils my intention.

For the second exploration with my rig I wanted to demonstrate how the rig could be used in a game setting. I thought that the capabilities of the rig could be demonstrated by making a combat animation that could be implemented into a game. I didn’t want to use another person’s model in this instance and also didn’t want to waste time creating another asset so I decided to take inspiration from western MMA/boxing for reference that would allow me to use just my rig for the animation. Below is the video I took reference from.

I liked making a slightly longer cycle from the previous experiment with my run cycle. I think that this animation proves that the outcome of the rig is functional and able to be used for animation, which is the goal of my project. There are instances where elements of the rig clip, and I think with a longer amount of time to weight paint in greater detail this would be able to be fixed.

after completing the run cycle, I realised that I wanted to show off just the sculpt and physical form of the rig, as well as find a way to introduce the rig in my final outcome video. during this week I created both the entering animation, which used the knowledge I gained during the weight shift practice in fundamentals. I had my character step forward, which will be him stepping into frame in the final outcome video. I am planning on having a sweeping pan that is close up, allowing viewers to get a full visual of the rig up close, since most of my animations will either be zoomed out fully or focussing largely on the upper body. For the same reason I wanted a long, fairly simple animation that would allow me to rotate the rig slowly on a turntable that would allow viewers to fully see the rig from all angles. As a result, both of these animations are quite simple but serve their purposes towards my final outcome.

in addition to both of these animations, I created facial controllers for several mouth expressions, such as smiling, pouting, inflated cheeks and kissing. this is so that in the final outcome I can demonstrate the facial capabilities of this rig. Creating these controllers is a little time consuming but otherwise simple- by adding a boolean attribute to the custom controllers I made, I can make keyable sliders for joint positions that can be set through the driven key option in maya. This is also how I created the controls for the hands and feet.

After finally being able to move on to actually animating for this project, I decided to start with something simple.

The first animation that I wanted to make with my completed rig is a run cycle, since it requires full body movement on both sides, and would clearly demonstrate if there were any major issues in the most basic parts of the rig. I was pleased to find that this animation went smoothly and worked well. I was able to use my custom controllers effectively to create this animation, including the reverse foot controls that I set up, and the hand controllers that I created for ease of use. Throughout this process there were some minor tweaks that were highlighted as essential to fix, such as some small areas that were missed in my weight painting, but these were easily resolved.

After creating the joint based rig I wanted to create custom controllers for each element of the reg. I used nurb curves to parent each joint that could be used as animation controllers. I also created a reverse foot control for the ik legs that would allow for more specific types of movement such as the character being able to come up on their tiptoes and stub the ground with the ball of their foot. I followed a tutorial to create this reverse foot control which I will share below: 

As well as creating the reverse foot control. I also wanted to create some set driven keys that would allow me to quickly animate some complicated parts of the rig, such as the eyes for blinking, the hands, and the face. I created custom controllers that I could set driven keys to which would allow me to animate much more quickly and easily in the limited time I had for the project. Having experience in animation, I was able to understand which controls would help me animate more efficiently in the later stages of this project.

I decided to create a custom rig for this character since I wanted to implement facial controls and create custom controllers. I have some experience in rigging, and as a result I found that I enjoyed this process. Once again on this project however,  I found that if I had more time there would be things I would add. I would have like to create switches that would allow myself to switch between IK and FK on both the arms and legs of this character. As a necessity I decided that for the sake of the project I could manage with IK legs and FK arms, but if I were to be using this rig for any extended length of time these would be the main features that I would be implementing to this rig.

In terms of topology, my limited knowledge definitely has an impact on the weight painting. In areas where I made sure to follow standardised good topology I found I was able to very easily weight paint, such as in the face area, the arms and the legs. However, in places such as the fingers or around the fairly complicated shoes, I found that the way in which I had topologies negatively impacted the way that I weight painted my model. I found this as a good lesson for a greater understanding of how all of these processes impact each other. Good weight painting is completely essential for good animation, and while I took care to complete this process carefully, there are certainly still parts of my rig that could be improved with finer tuning of the weights.

I struggled with this stage of the processes somewhat for a multitude of reasons, largely in that I found that the mistakes I had made while creating my lowpoly had adverse effects on the ease of which I could lowpoly. Not only this but I found that my model had so many pieces that I needed to spread my UVs across two different UV sets. This was quite a convoluted process but eventually I was successful enough to be able to create basic textures for the model. During this process I also tried to create normal maps to better indicate the detail I had been able to create in while in zbrush.

Ultimately I was not able to create a result I was happy with in the time that I had available, so decided to proceed without. This is something that on reflection I would like to have changed about my project- I think that the quality of the final outcome would be considerably improved by the implementation of normal maps, and if I were to return to this project in the future I would like to complete this step. However, for the sake of completing the project, I decided to move on from this point with a simplified smooth sculpt of my character. I decided to simplify some of the details on the clothing and body, which work better without the normal maps, which as a result has led to some differences between the final model outcome and the initial plan for this character.