Two Characters and a Setting Project

Brainstorm and notes on how two characters could interact within a setting

To begin this project I prepared several ideas on a sheet of paper to help brainstorm how I could tackle it. I started looking a how the two characters could interact with each other and what they could be before looking at the setting, as the two characters would indicate the setting they would be in (e.g. spaceman + alien = space/alien planet) I then made 3 sentences which described 3 ideas which I thought would be the most interesting to animate.

I then prepared two ideas for the pitch- the Swan story and the Plant story. Both started out very basic, and it took some time to think about what the actual stories would be. I knew I wanted to work with Maya so I thought it would be best to keep the character designs and animation movements simple as leave time to design and rig the characters/backgrounds, as I would be learning as I went along. It also would allow me to make the movements seem more 'real', and not rushed if I could take my time perfecting them.

My first idea, about a swan falling in love with a pool floatie shaped like a swan, started out as a simple sentence. I had a couple different ideas regarding how he would react to the float, and what he'd do to try and win its heart, including the introduction of different elements:
1. instead of the fake swan being a character, an observing pigeon would be, and the fake swan would be a prop
2. the inclusion of another floatie that acts purely as a prop to drive the story

I decided to go for the second idea as it would make the story simpler without the inclusion of a character that would be a lot more minor compared to the swan, as they would have very little interaction with each other. Furthermore it would be easier to model and rig a character that is very similar to the floatie, therefore saving time on the character-creating step and leaving more time for animation.

Script and early character concepts for the swan idea

Plant script and final plants, dropped because the story wasn't as strong as the swan idea

I then experimented with some initial ideas to what the characters and setting would look like. Knowing the limitations of my own skills with Maya, I wanted to keep the characters simple and cartoon-y, but also be obvious enough to what they were. Over the Christmas break I experimented with creating other objects, such as a tree and the original floatie swan (which inspired the film). I've found that the tree is stylised but nice enough to be able to be used in the film if I need to create a more interesting/detailed setting.

A selection of objects that I had practiced to create over the course of late Semester 1 and through the Christmas break (Girl, Swan floatie, Tree)

From the character comedy workshop with Alan, I took ideas on how the characters and setting interacted, as well as how the story progresses from the beginning, the events in the middle, and to the end. From this I created more ideas of how I could end the film, and by asking other people on the course, and friends, I created a number of small ideas how it could end.

Notes from Alan's comedy workshop

After sorting out the story I created a storyboard to help visualise what it could look like. The character designs were loose and not the final ones, as I was in the process of deciding my style. Other than the fact I wanted it to be in 3D, I had not given it much though on the film's aesthetic, other than drawings and sketches of the characters. Once I had the story down from the first crit, and Alan seemed happy with what I had prepared, I then created the characters. I had practiced with 3D modelling over the Christmas holidays and the previous semester, but I had not textured or rigged anything before, so this was a very exciting learning experience for me.

The storyboard I pitched to the group at the first meeting

One aesthetic decision I made whilst designing the characters is to incorporate a "low-poly" style- a simple low polygonal model that keeps things simple by directing focus onto the textures placed on top. This stylistic choice means that the characters could be kept simpler, e.g. less feathers to model on the swan, and therefore meant that rigging would take less time and would be easier for a beginner such as myself. Personally, I also believe that simple 3D rendering is pleasing to look at, and has been an effective choice in many 3D animated films and video games, such as Elizabeth Ito's "Welcome To My Life" and "The Legend of Zelda: The Wind Waker".

I experimented with a number of face/beak shapes, looking at real life proportions of swans whilst visiting a pond at Holyrood Park. I recorded them both through life drawing and photography in order to understand how they moved their necks and bodies while they fly or swim.

Primary and secondary resources of research that I used to observe how real swans move

Drawings of real swans from life at Holyrood Park, and drawing from a photo of the original pool floatie that inspired me

Development drawings of the swans and flamingo

Facial expression development for the swan, and concept art for the plant idea

After finalising a general design (but with room for improvisation in regards to the facial structure) I began to put together the actual characters for my film. I began with reference images, in order to keep proportions correct through the front, sides, and top view of the characters. I then blocked out a character from that reference on Maya. I began with the floatie swan as I could use my practice dummy as a rough idea for what I was going for, but wanted to push myself and make it more appealing in regards to colouring- and to give the inanimate object a 'life' and character of its own, while still retaining the fact that it is a fake both in the story and in reality. 

Final sketch for the swan, affectionately named 'Derek'

Derek's top, side, front, and back references for Maya

Shots demonstrating the references in Maya

WIP shot of a side profile of the swan- blocking out neck and torso

Final model pre-textured and without eyes

Final model ready to be rigged

Close up: Eyes are rendered to be shiny compared to the cartoonish textures of the swan's body 

Floatie swan's references

References in a 3D space

WIP of the swan being made within the confines of the proportions set out by the references

Final model pre-texture

Texturing the model with a UV map in order to understand how UV works- Note how the mesh is actually two objects: the body and the neck/head. This is because I was not going to put any detail on the body- just a plain white plane- I could maximise space on the UV for the face and beak.

Side profile shot of the finished model, with X-Ray visibility on to emphasise the tube inside it (like a swimmer's rubber ring)

Same side profile shot, but with incandescence rendering on- the 'cel-shaded' look I am going for, but is not good at showing the shape

Floatie flamingo's references

Floatie's references in Maya, demonstrating how I'd keep the model in proportion to its concept

Rough blocking out of Floatie

Rough blocking out of Floatie in regards to its references- still missing its wings and tail

Final model of Floatie

Final model with texture and smooth preview, which was used to emphasise the fact that its a blow-up pool toy by reducing its blockiness

I aimed to make the flamingo more masculine in a sort of non-conventional sense, as it would be the pool swan's assumed boyfriend by the real swan (aka Derek), and therefore a competitive sort of rival. I'm unsure if that was portrayed in the final model though, but I do like how his smooth model turned out as I had trouble getting the beak right in the character art I had created earlier. 

After I had completed all the character models and their textures, I moved onto rigging. Like texturing, I have no previous experience with rigging, and so had to learn with Mike. Fortunately only Derek needed an accurate rig and thanks to the simplicity of the model rigging and painting the weights was not much of a problem. Since Derek's main movement comes from his neck, I gave it a few extra joints in comparison to his wings or body as it was his main point of emotion, considering the fact that I was unsure whether I should implement blend shapes in the eyes at this point, as it could be considered distracting if it looked 'unclean' or if they eyeballs clashed with the mesh of the body. I tried to rig the controllers similarly to how Norman was rigged, as it was simple, concise, and easy to use.

I considered implementing IK/FK rigs into this model, but thought against it- it would not be worth the time as Derek's wings do not act like human's arms and so do not bend the same way, and for the film it would be easier to manually position them as they swap from acting like wings when he flies to acting like arms/hands when he holds the flamingo. Furthermore there wouldn't be much use putting them on the legs as Derek would be swimming for the majority of the film and IK/FK rigs would not add much benefit to his free-floating legs' animation quality. With the neck, I wanted to be as free as possible with it so I could squash/stretch it and bend it to emphasise its springy nature, I felt that using IK/FK would be a limitation on that freedom.

Rigging Derek

Side view demonstrating post paint weights, showing the neck working correctly.

All the controllers for Derek's body and eyes, created with NURBS circles

The rig, organised and labelled into different parts of the body

The floatie swan did not need any rig, so was able to create a simple floating animation test to see how it would look in the water. After getting used to rigging with Derek, I created a rig that was not as detailed as the first one for the flamingo, as in some shots his model distorts. Using a rig would help do that, and I was able to squash and move it to make it look more flat. 

UV Map of the fake swan swan (model seen above)

Original concept sketch of the pool, not the busy hills and city 

The background and setting was always going to be a pool, but I was unsure whether or not I should add a more detailed background or not. I drew multiple different sketches to try and solve this problem, but found that adding a more detailed backgrounds became distracting to the main attraction. I tried to match the low-poly aesthetic by practicing with Adobe Illustrator to create vectors of palm trees which could be used to texture planes and placed around the scene, but found that during 10x10 I became interested in the background that was created with masks and solids in After Effects. The gradient helped to create a spot-light effect on the pool, therefore enhancing the audiences' focus on the point of interest- the animation itself. After asking friends, I decided to drop the original idea of having things in the background and just keep it simple. 

 Coloured in sketch, in which this was the point where I noticed the background may not fit with the polygonal style

Concept of using vector images as background pieces created in Illustrator

I used 10x10x17 as an opportunity to practice moving both models and the swan's rig. I created the film using the same background and basic story in order to visualise how the film would look aesthetically. 10x10x17 also allowed me to get more comfortable with the use of the graph editor in order to create more natural movements, such as bobbing in the water, or to create a continuous cycle or animation, such as the swan's legs moving on the deck chair.

During Michael's After Effects introduction class, I used this opportunity to animate Derek's thought bubble scene purely in 2D, in order to emphasise his imagination, and also allow myself to diversify and try new things. All the assets were drawn in photoshop, and put together in the workshop.

I used my storyboard as a checklist and guide as I animated the scenes, starting with the shots that have gentle movement, such as the beginning scene. I allowed myself to become more comfortable with the software, practicing more with the graph editor in order to smoothen out the movements to portray the idea that the fake swan was being pushed around by a gentle breeze.

Example from scene 21 of the fake swan's graph editor

Above the data shows the movement of the fake swan in scene 21. The data at Rotate X and Rotate Y (centre red and blue lines) create its 'bobbing' effect, and you can see the point in which the swan makes contact with Derek as the Translate X (red line on top) data creates a downwards curve instead of an upward curve. The Rotate Y curve also takes a sharp turn as it shows that the swan has collided thus changing its direction from being constant to a sharp downwards arc.  

Some animations, such as the shot in which Derek slaps the flamingo, were cut from the final film, due to the fact that I struggled to get the movements rig. However whilst trying to animate I was able to use the grease pencil tool, a feature I didn't know existed. Whether it was a problem with my rig, or the animation, I am unsure of, but something didn't feel right, and so I thought it was best to remove it.

Putting the film together was hectic, as I chose to use the renderer mental ray- which in all honesty did drive live up to its name. Choosing to set my materials on my characters as incandescence reduced the shading on the models, which helped emulate the 'cel-shaded' look I was going for. However the time that it took to render all these images out was longer than expected, so I had to go through each shot and clean up by removing unused assets such as references, props/characters not in the scene, optimising the scene size, and deleting non-deformer history.

After rendering out everything, I added the background layers in After Effects and exported the scenes to Premier in order to put the film together. Without the aid of music or sound, I found it difficult to get the timings right, and I feel some of the humour is lost with the loss of sound. As a post-hand in improvement I aim to put together a soundtrack with one of the composers from the 10x10x17 project. So far there are a few people interested, and I am curious to see how each of them interprets the film.

Overall this project was the highlight of the year, as it allowed me to truly create a story and characters without limitations. I have learnt so much from this project, and though it could be argued that it was very ambitious, I believe that there is a lot to show in terms of story and character development. In the future I will definitely use Maya again, but also enjoy how the 2D and 3D segments compliment each other, and might try to combine stop motion and 3D again like in the World in a Room project.