DEVELOPMENT & PROTOTYPING
Updated: Mar 18, 2022
TASK ONE: PERSONALITY WALK
I started with a personality walk from a shared and provided reference for the whole class. I initially had difficulty with setting up the scene, specifically the Image Plane attached to the Camera, but I found it challenging in a good way. I found it tough to set up the scene initially, but once I made note of how to do it, I assume it will improve on my pace in the future.
After Splining and Polishing, I found the character to slowly move into each pose rather than fluidly copying the movement in the reference. After some troubleshooting, I realised this was because the movements would Slow In and Slow Out instead of linear. After making certain tangents linear and others splined, the movement became a lot more believable in its speed. An error I made on my part however was the up and down movement of the character in their walk. The movement caused a serious level of disbelief, and I found this difficult to correct after splining.
I decided to move forward onto my Personality Walk. I recorded a reference myself, but after some discussion with my tutor, I found it was bad because it was a creeping movement and not a walk, so after analysis I chose an exaggerated angry walk, which was greenlit by my tutor.
SyncSketch was a fairly new format for me to use for drawing the outline of the poses. In order to properly capture the up and down movement, I used Photoshop then went to SyncSketch again. I found it difficult to use my mouse or trackpad to outline the edges of the silhouette to create the pose, so I attempted with my Wacom Tablet. Using the tablet, outlining became a far more smooth process. As a beginning artist, my drawing skills were weak, so the outlines weren’t smooth or confident as I saw in the provided reference. Yet, they were serviceable and helped with my final animation.
(Above is a progression reel of some of the many playblasts I made for the Walk Animation.)
TASK TWO: PERSONALITY PUSH
After completing my Personality Walk, I moved forward to the Push Reference. I began using the reference provided to us by our tutor. The first error I made was assuming this process would be similar to the personality walk.
There were a number of differences with it: firstly, the character does not move in Global Space, but rather in Local Space to the cupboard. The character’s movements are defined by the movement of the cupboard. If the cupboard moves forward, the character’s body moves forward while their feet remain planted in the ground with little to no movement.
The most challenging parts of the personality push were the foot placement and the hand positions. The foot placement became difficult due to the actor in the reference provided pushing a wall rather than a moveable object. My tutor gave some advice on how to get past this: the feet placement remain mostly the same to each other when the pose changes as the feet slide back away from the wall; so instead of moving the character backward, have the character’s feet remain in the same place and their upper body push forwards with the cupboard. Although this piece of advice was helpful, I did struggle with grasping how to apply it since it was such a new concept for me.
The second challenge was the hand positions on the cupboard: since the hands would touch the cupboard, I found it very challenging to have the model touch the cupboard without the rig’s hands phase through the object. After some struggle, I realised it didn’t have to be, since Keith Osborn in his book “Cartoon Character Animation for Maya” argued that as long as the audience will see a pose that is well done, the poses you don’t see can end up Breaking Joints. If the audience can see the hand coming in contact with the cupboard and not notice the slight millimetre of the hand in the cupboard, it would not matter.
I began by taking a reference of myself pushing a wall in an exaggerated manner. I couldn’t find any large objects to move at the risk of breaking them, so I settled on utilising the same object the previously mentioned reference had used: a wall. I wanted to showcase in my character’s movement the weight of the object and how the character utilises their own mass to match the mass of the cupboard.
I took the footage directly into SyncSketch. (I would not need to use PhotoShop since there is no up-and-down movement of the character to mark.) So I took every pose that had a major shift and outlined the character in the reference, and I outlined each pose in-between I realised could potentially be important. If my actor raised their foot, I would outline the foot before the raise and the foot coming in contact with the ground. To make sure this movement was captured, I outlined the in-between where the actor’s foot was raised high before falling down.
(Above is a progression reel of some of the many playblasts I made for the Push Animation.)
I was really happy with how the outlines came out in comparison to the Personality Walk Animation since I grew more confident with my strokes and they became more defined and proper.
TASK THREE: LION WALK ANIMATION
This was the most challenging of the tasks I faced. It was a Cat Animation with a performance piece and walk.
I found some clips as reference and used them for my animation. Based on my peers, they decided to do a lying down pose to stand and walk, but I wanted to experiment with a different type of performance. Searching through references, I found a reference where a tiger jumps down from a platform and walks to another side. In addition to the clip, I used After Effects to combine the jump down clip with a clip of a male lion walking. After outlining the first clip of the tiger jumping, I began outlining the male lion clip and I encountered an issue.
The male lion’s mane was too large and covered most of the body, so I was unable to accurately see the movements of the lion. So I had to replace the footage for the walking cycle. I searched and found a tiger reference to fit the Lion rig. I began outlining the Keys in red, and outlined the inbetweens in yellow.
Outlining the Jump took time, but based on the end result was necessary and worth the effort put in. I surprisingly finished the jump ahead of schedule, and needed to move forward. So I attempted to continue forward with the animation based on references without outlines on it. I was fairly surprised to realise the outlines were hindering some of my poses instead of helping. I found my poses were clearer and defined when doing the non-outlined sections of the clip rather than the outlined.
I decided to take a Straight Ahead approach to the lion animation rather than the Pose to Pose approach I tried with Tasks 1 and 2. I decided to do this based on the perspective Keith Osborn gave in his book “Cartoon Character Animation in Maya,” where he argues Straight Ahead animation gives a more unpredictable feeling and draws attention to solid animation which is detached from the stiffness of Pose to Pose animation.
I set up my monitor with a four-panel layout: one on my final render camera, one for selecting NURB curves, and the final two for both orthographic sides of the lion. I found this quite helpful since I could focus on the sides of the lion as shown in the playblast, as well as notice any mistakes in the posing when viewing the final render camera. I would encounter an issue where Maya would slow due to a wide range of tasks it was accomplishing and therefore I couldn’t readily see my work, but my tutor informed me of a simple solution to playblast my work to view it each time.
I began my work by focusing on the poses of the lion, and the detailing later on. When I first began in Stepped, I found it took a long time to do the poses for the jump since it was a tiger instead of a lion. I found their body sizes actually affected their movements: where a tiger could take long strides and have a greater impression of weight due to having a greater mass, whereas the lion would not. If I had come to this realisation initially, I may have done things differently.
After posing, I focused on the details. Such as the spine, head, tail and shoulders. I found the shoulders quite challenging, but I ended up enjoying splining and polishing them since the satisfaction of the completed and well done animation had been achieved.
(Above is a partial progression reel of some of the many playblasts I made for the Lion Animation.)
Overall, I really like my Lion Animation, but if anything were to change I would definitely polish further the walk itself. I just find that a work that is completed but imperfect is better as opposed to a work that is consistently polished and incomplete.
TASK FOUR: UNREAL ENGINE
The UnReal Engine Task was fairly interesting, but I encountered a few bumps along the way.
I remain mostly unfamiliar with UnReal Engine as opposed to Maya, so I found the smallest of details the hardest to work through. Finding a certain button, how to link two sections together, etc.
Task 4 was to export our lion animation from Task 3 into UnReal Engine, and have the lion move in a global space in an environment. I found it easy to transfer the Lion from Maya to UnReal using Cache and [name] to export then import to UnReal.
I quickly encountered an issue. My lion animation was specifically of a Lion jumping from a height onto a flat surface, and continuously walking. The environment had a few surfaces but a number of them were rocky and uneven to give the illusion of a proper environment. I added an asset of a rock at a certain height to create a new level for the lion to walk.
I added a fair amount of foliage to the task, and I found it fascinating how UnReal allows to “paint” assets onto a surface.
I mostly was able to achieve the final end result, but I wasn’t entirely happy, so I redid the keys for the camera a few times and ended with my final result. I was mostly unhappy with the composition and how the camera didn't render the lion the way I wanted to. I planned to have the lion jump down from a height and crawl through certain foliage to act as it it were hunting. I did this by subverting conventional composition and had the lion avoid lining up with Rule of Thirds or the Golden Ratio.
(Above is a progression reel of my renders I made for Task 4, but was ultimately unsatisfied with.)
Overall, I’m satisfied with the end result, but I feel like more time to familiarise with UnReal Engine and its basics would be necessary for any future work and tasks.
Comments