Categories
Virtual Spaces

Blog

For this assignment I am making one character in a game for the theme of cute festivals of my culture. This character is targeted to a young audience. Therefore, I need to draw/sketch how the character will look like. I first need to plan out the steps involved in making this character and then create a mind map of how this character will look like and what themes I am considering. 

  • Character Design 
  • Character 3D Modelling 
  • Character UV & Texture 
  • Character Rigging 
  • Character Animation & MoCap 
  • Character Rendering 

Since I am making a MoCap to create the animations of the character, it must have 2 legs, 2 arms, a head and a body connecting all limbs. A humanoid rig. 

Picto Plasma – similar style 
Cultures? Vietnam  
Festivals: 
Tet à Red Packets, Lion Dance, Zodiac, Flowers, gold 
Hung King à Temples, Dragon, King

During my research of this mood board I started to settle with going for a Chinese humanoid lion warrior mainly because the limbs are easier to animate through mo-cap.  My first thoughts of the design of this lion was going to be in the style of Alex from Madagascar but have facial characteristics of the lion costume and statue as well as the armour of a Chinese warrior. Using my mood board I will design my character. I felt it was a good idea to create a character profile to cement my vision of this character will look like. I had to also think about the colours and basic textures of what the warmer colour pallet –> analogous colour scheme –> groups of three colours that are next to each other on the colour wheel. I did want vibrant colours because it would appeal more to children. 

Character Profile 
Name: XIAO Zhan 
Gender: Male –> because of mane  
Cultural Background: China 
Symbolization: Lion à strength, stability and superiority Dragon à power, boldness and excellence 
Purpose: educate children the significance of family, friends and the sense of community. He’s seen as a hero and a symbol of unity. 
Education: Gifted 
Fears: none –> built different 
Religion: Buddhist

3D Modelling the Character
I first started out the lion using a pre-made model on Maya Autodesk. The model was titled as monster figure and the shape was somewhat humanoid but also very vague as to what the character could be. I found this to be time efficient as then I wouldn’t need to spend time blocking out the model and can go straight into the details of the lion. During this I realized there were some parts of the lion that were going to be props that would be made separately. This was because of how the rigging system works on Maya. Therefore, the mane, sword and skirt will be modeled, and UV wrapped separately. During this process I took a very long time to feel satisfied with the outcome of my models. It was frustrating since I knew that without the model being finished I cannot UV wrap to then texture my model and I cannot rig the model.

Lion with eyeballs

– Skirt 
This was very simple to make and straight forward but the only thing I wanted to do was make the skirt flow using Maya’s physics plugins, but I was advised not to do this as it would take more time and could provide more complications to the rigging process.  
– Sword 
Modelling a sword was harder than expected I thought the simple geometry of a sword and the multiple tutorials online on how to make one would make this straight forward but when I was making it, the edge loops would be misaligned to where I wanted them, sometimes the geometry was buggy and hard to shape. Eventually I did make it, but I sacrificed the extra details on the handle to save time since it was only a prop and not too significant to the project. 
– Mane 
At first, I was going to create a sort of helmet and the use sculpting tools to make it look like flowing hair but then it didn’t look appealing aesthetically, so I created large follicles of hair and copy pasted them over the scalp of the lion’s head. This took some time to position the follicles as there were many and angling them to resemble a mane of a lion was difficult. In this situation I was focused on making something that looked appealing more than accurate to a real lion. 

UV unwrapping and Texturing 
Once I finished modelling, I would UV unwrap each object through camera-based view and then cutting along edges of the object depending on how it would be textured. The skirt I just cut in half from the back side, the sword along the middle so it was sliced in half vertically, body of the lion was split through the major limbs and the mane took the most time as there were many follicles.

After UV unwrapping, I was able to begin texturizing and colouring the UVs. I downloaded the UV maps and went onto adobe photoshop to paint the UV. It took some time as I also needed to refer to the UV editor to add details onto the UV for example when painting the belt, I needed to know the centre of the belt is and how big the buckle would look like. I also would paint over lines to ensure that the entire UV would be filled with that colour. I also needed to make sure that it didn’t cross over into a different UV. In other words, colouring between the lines. This was straining my hand a lot because I had to be accurate with the painting. I then needed to select materials that the UVs would be put on. I mostly played around with the reflectivity and roughness of the materials so that it would look like clothes and less like shiny armour.  

Rigging 
The quick rig step-by-step feature made this project much easier than the longer method of rigging. The problem lies where I initially made the model not in T-pose. Apparently, this creates problems when skinning the rig to the model. To fix this I created a rig onto the non-t-pose lion so that I can move the arms and legs so I can put it in t-pose. I then deleted its history and then removed the control rig and then redo the control rig again. I now do the weight painting. Placing the guides for the rig was simple as you just needed to move them a little bit to where the joints would be on your model. To make sure they are positioned correctly I use the plane views to help me with this. I did initially have trouble with it because Maya would start putting influence into unrelated parts that I cannot erase easily. It is hard to explain but there would be spikes coming out from different parts of the body. The fix for it was instead of using the black and white paint I used the multi coloured one so I can see more clearly the levels of influence and can cut them out. I would test the rig by moving the joints and seeing if it moves smoothly or not. If it didn’t then I go back into weight painting to adjust the influence.

Animation 
Since there was no mocap available to me I used animations on Mixamo.com. I picked an animation that wielded a sword since my character has a sword. Surely enough there was an entire pack of them. I uploaded the character to Mixamo and by luck it accepted the rig I put on it, so I then downloaded the entire pack. The next part was a little harder and time consuming, I now needed to clip all the animations together. For each animation clip I used the Trax editor and I had to define the frames and slided along the order of animation. I wanted the animations to blend through each animation but I wasn’t able to do that mainly because they were pre-made and were not made to be in sequence. Another problem I had was the positions of the character when I would play the animation. Since the animation has its own position co-ordinates I had to modify them. Fortunately all I needed to change was the reference position so I set them to y=0 (which is the ground for my project) However, I think there was some complications to it. This works only for animations that aren’t jumping or crouching, for those animations I would need to offset the positions so that the lowest point was at y=0. I created a camera and aim so the camera was static in one position and it would aim at the center of the lion. I then created an environment similar to a studio shoot with area lighting from top and both sides.

Trax Editor

Rendering

Environment w/ lights, camera and background
Aiming camera onto character rig
First frame of render
Categories
Virtual Spaces

Critical Appraisal

I found this task to make my character to be very time consuming as a process of completion. I often felt like I had wasted a lot of time on something that ended up either not working, wasn’t important, or errors with the software or files. Nevertheless, I found this project to be quite eye opening and taught me to be more objective focused but also sparked my creativity in the process. 

I was very happy with how the texturing and materials turned out except for the hair. I forgot to make the hair less shiny, so it didn’t look like metal. It was time consuming and took a lot of patience to paint the UVs the way I wanted them, but it was worth it at the end because they turned out quite appealing. The UV mapping was also very streamlined and didn’t cause me any problems and I think that was because I struggled with this step before in the other projects but now, I became comfortable and aware of obstacles that came with it. 

I wanted to wear a mocap suit to create my own animations but however it was expensive to rent a suit and unfortunately even if I was able to, I later found out that the file format of the mocap would’ve been incompatible with Maya and therefore a waste of time. Therefore, I decided to use pre-existing animations because it would be faster, its free, easily accessible and judging from my abilities to act a better-quality animation. During this project there were multiple times I had to redo the rigging of my character for the simple reason that sometimes the rig would de-skin itself from the model and therefore I had to remove the rig and start setting up the skeleton, control rig and weight painting. This was utmost infuriating especially when it happened right when I was applying the animation to the character. This was unavoidable so the only measure I can think of is to always save my work often especially before performing something that heavily alters my work and strive to work ahead of the expected progress so that if there are problems, I am not set back by them in the long run. 

Unfortunately, I heavily misjudged how long it would take to render. I did a relatively high 60fps animation. I tried to offset the time to render by reducing the resolution to 540×960, overclocking my CPU, keeping my room as cold as possible, pointing a desk fan at my laptop and elevating the laptop to let more ventilation to cool down in hopes that my laptop could render out the images faster. However, I felt that if I do have an animation of my character I would have succeeded as I personally would know that if I had more time and/or rent a more powerful PC I would’ve produced a longer and higher resolution animation. Regardless I felt that I was still able to show my knowledge and skills in creating a simple character even with a shorter and lower resolution animation. 

Overall, I learnt during this project that even if I knew the steps to do something I shouldn’t disregard the possibility that it doesn’t go the way I planned. Countless of times I felt that I knew what I was doing and therefore I wasn’t too worried about the deadline, but I should’ve worried about how long it would take to do all the steps. During class and workshops, I would focus on something that I wasn’t sure and ask questions on how to continue, what was possible with the project and what I should prioritize on. 

I was not confident in my abilities design a character as I’ve had little experience in drawing creatively. This didn’t help understand how I can envision the character because I wasn’t satisfied with how I designed it. Furthermore, I was too focused on the context of what the character had to be that I completely disregarded how hard it would’ve been to create the design I made. I was initially planning on making armour that would be put onto of the lion itself, having fluid objects such as the hair and clothes and being rendered using Maya’s physics plug-ins, having rigged fingers and eyes. Eventually in this project I found out there were many complications to this and had to resort to more simpler methods to achieve a similar effect.