Animating the Owl

I animated the owl for all of my scenes in Maya as did the rest of the group. The first thing I did was analysed some reference footage of owls flying on Youtube to base my animation off. I noticed how the wings bend as they lift up and stay completely extended as they come back down. Luckily our rig was really easy to use for making the owl fly and it allowed us to make subtle changes to the animation. Sophie designed the owl model and rigged it and did an excellent job because the rig was very easy to use and make convincing animation with.

I set out each scene to be the exact duration as seen in the animatic (this is why having a frame counter on the animatic was very useful). I started animating each flying shot by doing a simple A to B with the rig to block out the timings. I used the animation layers in Maya for all of my animation. This was a workflow I learnt from the Youtuber Sir Wade Neistadt. On one layer I put all the wing animation, this was so that I could easily manipulate the wings on their own without altering the rest of the animation. I got the animation right for one wing and just copied all of the values over to the other wing, except for the X value which I always had to invert. I also created another layer for animating lift on the owl, this meant that as the wings went down, the owl’s body was pushed slightly upwards and vice versa. Animating on layers allowed me to control the influence each layer has on the final animation with a slider. So if I felt that the owl was lifting up and down too much, I could just pull down the influence without affecting the graph editor.

This video is a much longer version of the video above. I created it as guide for animating the owl for the rest of the group. So this video acted as part of our production bible.

There were some moments in the story that were difficult to animate with this rig because of it’s limitations. These were mainly times when the owl has to land and take off. We tried creating another rig to do that, but we were running behind schedule. In the end I decided to alter the storyboard for those scenes to kind of cheat the owl landing. Some of the options were to have the owl be obscured by an object, have a cut away to a wide shot or a close up, but basically just have the owl land out of shot. But I think we got away with it, especially with the landing shot towards the end of the film.

Making the Soundtrack

At the start of of November we had a networking event with some music students to pitch ideas to each other and form connections with them. During this event we listened to the musicians tell us their specialities and strengths. I had a look through the online portfolios that were provided. After looking through all of the portfolios I quickly found a musician who seemed perfect for our project, so I quickly emailed him and he replied promptly agreeing to come onboard with our project.

Throughout school I had music lessons learning to play the clarinet and piano, and I played in some orchestras. Having a background in music allowed me to better communicate what we wanted with the composer as I knew some of the terminology to be able to form a common language. For instance, at about 1:50 into the animation, the owl travels through the dark smoggy sky, then the clouds part and he sees his bright home island. I wanted this part to be a big emotional moment, so one of the first things we added was a crescendo in that part. Then we played around with the texture and tempo in the bars before that to get a real change of weight in the music.

Once I had the animatic blocked out I sent him it to use as a guide for timings. Over the weeks we sent many versions back and forth as he made changes based on the groups feedback, and I altered the animatic to better accommodate the music. Eventually we ended up with an amazing soundtrack that perfectly fit the mood of our film and had all of the emotional beats we wanted.

Storyboarding

During this project I really got into storyboarding which I dabbled in a bit last year. Since I was the only member of our group who really had any experience with Storyboard Pro, we agreed to let me do the majority of the storyboarding.

We used Microsoft whiteboard to thumbnail and brainstorm ideas. This was a place we could all log into and add images, videos, and drawings to. It was here where we came up with the idea for our story. We decided to assign each island in the story to a different member of the group so each person could be responsible for their own island. We thumbnailed short sequences for what was going to happen when the Owl lands on each island and we came up with the idea of it collecting sticks. I then took all these thumbnail sketches and redrew them in Storyboard Pro and added in a few scenes to fill in the gaps.

I watched Dermot O’Connor’s LinkedIn learning course on storyboarding. I thought this was a really good in-depth course. It was really useful to revise shot types and film grammar which I applied when creating the storyboard. I also applied posing and staging rules such as the 180 degree rule.

I posed all the scenes I worked on very carefully because I was working to make the animatic watchable in two different aspect ratios. We created the original version with a square aspect ratio so it would look good on social media. We also created a 16:9 version to be viewed on a regular screen. This meant that I had to consciously compose each scene so that all of the important action occurs in the middle horizontal section of the square, this was so the top and bottom could be easily removed to make it 16:9.

After I finished storyboarding, I used Storyboard Pro’s timeline feature to create a basic animatic with rough timings in. I was then able to send this to the musician so that he could start working and make some basic changes to it. I made use of the 3D camera in Storyboard Pro and created some parallax effects in some of the scenes.

Modelling in Blender

Once we began designing all of the islands, we decided that each member of the group should model something each. So 4 of us modelled islands and Sophie modelled the Owl and rigged it. Personally I didn’t have much experience modelling in Maya before this project but I did have some experience modelling in Blender. So because I was able to offer technical assistance and teach the rest of the group the basics of blender, we decided to model all of the scenery in Blender then animate everything in Maya.

I started by modelling a proof of concept style frame for the first island based on a drawing I did. I created a basic icosphere then carved it out with the sculpt tool. I found my Huion graphics tablet incredibly useful for this as I could literally draw and carve out the island, and control my pen pressure. I played around with the settings of the brush in different areas to achieve different effects. It was important to make the island look low poly because that was the style we were going for. Luckily when you carve with the sculpt tool you can control how many subdivisions you are creating, so I kept that number low for most of it.

I found a really useful tutorial by CG Geek which was very relevant to our concept. In the video he goes through all the basics of modelling in low poly and touches on making a simple node network to create transparent water.

I went through a very similar workflow to create my actual scenes that ended up in the final animation. When each model was finished, I exported it as an Obj then imported it into Maya. Surprisingly, we didn’t have many issues moving the project back a forth between Blender and Maya. One of the main things was that textures didn’t transfer over easily. So I didn’t do any texturing in Blender, I just added basic lambert textures to the models in Maya. Some of the models required me to separate sections in blender and export the island in pieces and reassemble them in Maya to add the textures. For the wide shot with the large forest, I used MASH with a randomizer to spread all of the trees around a plane. This was a big reason why it was important to split the island into pieces, so that I could control where the trees would generate.

Remote working

One of the biggest challenges this year has been working remotely. This is something that makes group work especially challenging since communication is a lot more difficult. It was essential that we created some sort of communication channel that we could all use to talk and send files. So we decided to use Discord.

Discord is a streaming platform made for gamers that allows you to text chat, video chat, send files, and share screens live. These features were all essential to us which is why we decided on using it. During Manchester Animation Festival I was surprised to learn that throughout the lockdowns in 2020, many animation studios used Discord as their platform of choice for communicating with employees. This surprised me because up until recently Discord was a platform used only by gamers but now it’s being used more professionally by companies. This is another example of how the gaming industry has played a big part in the progression of technology, like the advancement in graphics cards and raytracing technology. One of the reasons why a lot of studios were using Discord over alternatives like Zoom or Microsoft Teams, was that there is the ability to create meeting rooms that people can drop in and out of as they wish. With applications like Zoom, teams have to organise a time when everyone is free for a meeting and stick to that, but on Discord you can join a meeting room on your own and wait for others to join without an invite.

We organised our server by splitting up different parts of project into different chat channels. So we had one for preproduction, production, postproduction, and general things.

We briefly tested Discord before we started this project during the 24 Hour animation challenge.

A couple of us from the group decided we’d try to physically come into Uni everyday when possible to get the most out of the facilities and meet in person to better communicate and solve problems. However, with the unpredictability of the university access, this wasn’t a very reliable plan. Which is why it was so important to come up with a good virtual communication system.

24 Hour animation Jam

This October I took part in the 24 Hour Animation Jam with a group of 4 others on the course. I treat this as practice for the Creative Collaboration Module as it is essentially a very short version of it.

This was one of the hardest challenges I’ve ever done as staying up all night producing an animation creating everything from scratch is not easy. Everybody worked remotely so to communicate we used Discord which was an essential for working together. Using this was valuable because it allowed us to get experience sending files, sharing screens, and solving problems as a team remotely which are skills that are increasing in importance.

We assigned rough job roles and created a simple schedule to plan out the mini project. Inevitably, we fell behind our schedule and found it difficult sticking to it strictly, and we ended up sharing jobs and helping out with each others work.

Finishing this project, I had much more confidence in working on a short film and had a better understanding of using an animation pipeline in practice. But I mainly had excitement to work on our main project. If we could work together to make an animation of reasonable quality in 24 hours, then it’s exciting to think of what we could make in 10 weeks.

Dog flour sack animation

I decided that for one of my exercises I would try to attempt a flour sack animation. So I began by downloading a 3d flour sack rig and having a play with that. After spending many hours struggling with the rig I decided to try using it in Toon Boom instead.

I wanted to practice by making a simple emotion to emotion with the model and trying to maintain the volume.

I’ve been reading Acting for animators by Ed Hooks and I wanted to apply some of his theory to my work. The starting point for creating my performance was to use his formula for acting. “Acting is an action in pursuit of an objective while overcoming an obstacle”. So I started thinking of basic objectives and obstacles a flour sack could have and I decided to give the flour sack some animalistic qualities.

The premise for this story is, the flour sack is sleeping and is woken by a bouncing ball. The flour sack desperately wants now to catch this ball but the problem is the ball is moving too fast and is difficult to catch. The sack then leaps forwards and ultimately misses the ball.

The flour sack behaves very much like a dog in this performance so I observed by own dog and how he reacts for reference. The Objective the flour sack has is, he wants to catch the ball, the Obstacle is the fact the ball is moving. The Action the flour sack takes to catch the ball is to dive forward. The sack misses the ball so ultimately fails in his pursuit. The type of conflict present in this performance is conflict with character, as the bouncing ball behaves like another character teasing the flour sack by waking him up, and then by taunting him.

I played with the flour sack rig in Maya to try and pose it but I came into lots of difficulties using it resulting in the rig not doing what I wanted it to and doing strange things when I animated it. So I decided to change plan and animate the flour sack in 2D. I animated this emotion to emotion to see if I could maintain volume in 2D.

dav

I came up with the idea of posing the flour sack in Maya (seeing as that wasn’t too difficult), then using those stills to aid my drawing and keep me on model. This is a workflow that I would use again when animating in 2D as I found it really helpful for establishing poses quickly.

Slide animation

For this performance I was inspired by my younger brother playing on a slide. He was trying to struggle his way up the slide the wrong way without much luck. So I thought this would be a good basis for a short performance with a character struggling to climb up a steep slope to ultimately fall down again.

I had fun shooting my own reference for this animation which was difficult as I am too big for most children’s slides. I reviewed the footage in sync sketch which was a very useful tool. The main things I took from the reference that ended up in the final animation were the foot placements and timings, which were unusual walking uphill. As gravity has more influence, the walking pace accelerates to compensate the pull downwards.

I storyboarded the animation in storyboard pro and produced an animatic. I wanted the steepness of the slide to look intimidating so I played around with having the camera at a low angle. I found a free model for the slide online (which I used in the animation) and used it as a drawing guide for perspective in the storyboard along with the Stewart rig.

When I came to creating the animation I used the workflow Blocking, Spline, Polish, which I learnt from the Youtuber Sir Wade Neistadt. For blocking the animation I made all my poses in stepped mode on the frames I had marked out in my reference and my animatic. Once everything was blocked I created playblasts of every shot and used them as placeholders over the animatic.

When I came to do a spline pass I used the TweenMachine plugin which is a plugin I had never used before but now absolutely love. It allows you to create inbetween frames and control the influence the next and previous keyframes have on them with a simple slider. This was really useful when adjusting the easing of the head turn which would normally have taken ages because my graph editor was messy as lots of body parts were animated. but it ended up being very simple as all I had to do was set the inbetween to favour the previous keyframe. Once I was happy with the keyframes, I set the curves to spline mode (but I could have left them on stepped and continued inbetweening), and after some polishing the animation looked great.

Another tip I learnt was to animate on layers, and I now do this for all my animations. Animating things on different layers gives you more control over individual aspects of the characters performance. It’s also none destructive if you want to delete something later on. I now always animate my camera on a seperate layer and lock it so when I accidentally move around in the camera view with autokey turned on, it doesn’t mess it up. To quickly check how my keyframes were looking, I kept duplicating the layer with the blocking poses on, switched from stepped to spline, then went back and make changes. As a final touch I created a layer with the characters chest oscillating to make it look like he was breathing to give him a bit more life. Because this was on one layer, I could control the intensity with the slider to change how exaggerated his breathing was. This could switch him from passively breathing calmly, to panting aggressively, all with one control and 3 keyframes.

Shoelace Trip up Animation

This performance is an observation based on my own behaviour that I picked up on recently. I always tend to be quite clumsy and always trip up on things in the street which can be quite embarrassing. I noticed one time when I tripped up and fell on my face that my immediate reaction (after extending my arms to break my fall) was not to respond to the pain I felt, but to look and check that nobody had seen me. I found this fascinating and everyone I asked agreed that they’d had the same experience before. So it seems that it’s human behaviour that we fear, and are more worried about being embarrassed in public than feeling actual pain. After doing a bit of research I found that there have been many studies done that prove this to be true.

So since this is something common in us all I thought it might be humorous and relevant to animate a character in this same situation.

Here is the animatic I created for this performance. It begins with a side shot of the character’s feet already in motion, as Ed Hooks says “scenes begin in the middle, not at the beginning”. We see his shoelace is loose and as he walks it gets tangled and he falls to the floor. We then switch to a front shot to see him fall and get back up, and it’s revealed that there is a crowd of people behind him who all saw him. The character then looks around, embarrassed, then finally settles with his head down. As an audience we are meant to feel sympathy for the character.

The only thing I did differently from the animatic was that I changed the shot where he falls over. I changed it from a front shot to a 3/4 shot to avoid any foreshortening of limbs and to get more out of the performance.

I filmed my reference for this animation in my garden (please ignore my very confused dog, he’s very photogenic and needed to be involved but doesn’t affect the reference). The main piece of reference I used was from 0:25 – 0:30 for the final shot where the character sits up.

I analysed this footage in Syncsketch where I marked out the extreme poses, breakdowns and other important frames.

I began creating the first shot by blocking in the key poses of the walk cycle using the Animators Survival Kit as a guide for timings. I blocked in the key poses on stepped curves, then switched to spline when I was happy with the poses. For the shoelace I used a mix of 2D animation by importing the playblast into ToonBoom then manually drawing the shoelace into each frame. The end result I think looks quite convincing. If I were to do this again, I would probably use the same approach as I imagine animating a shoe lace in 3D with convincing gravity and drag would probably be more difficult. Looking back at the animation I think there needs to be more of a response from the character when he initially pulls on his lace, maybe some followthrough is missing.

I posed all the shots in stepped mode basing them all on my reference footage that I had open on a second monitor. Then when the timings were right I set the inbetween poses using the TweenMachine plugin which I’ve found really useful. This final shot where he sits up was a difficult one to animate because I wanted to convey the shock and high adrenaline of the character, so I had to find the right balance between quick jerky movements and slower tired movements to contrast. I also pushed the extreme poses to exaggerate his head looking around and used some subtle anticipation and followthrough. I made a separate animation layer and animated the character’s heavy breathing on it’s own where I could easily control the intensity of it. Animating on layers is a workflow that I’ve found really helpful.

For a final touch I added in some background characters to be laughing and pointing at the main character. Unfortunately due to time, I was not able to animate these characters so I made the decision instead to try and put them all in strong poses that communicate what they’re doing. Looking back, it’s probably a good job that I didn’t animate those characters because I think if they were animated, then there would be far too much movement happening on screen and the viewer would be confused as to where the focal point of interest is. But as still poses I think it works because it could be assumed that time has slowed down for the character in this embarrassing moment, or we could be inside the character’s head and his anxious mind has invented this situation.

Teapot Animation

For one of my performances I decided I wanted to use an inanimate object and give it a character. While I was brainstorming and thumbnailing potential objects I could use such as a bottle, a chair, a mug, I was drinking a cup of tea, and I decided that a kettle/teapot would be a very good idea. So I began by thumbnailing some character designs. To me, kettles seem like very angry objects as they’re literally boiling and sweating, and traditionally old fashioned ones make loud screams when they’ve boiled. These ideas seemed like a good starting point to base my character and story on.

I created a very simple storyboard and animatic on storyboard pro. The story is made up of two very simple scenes. It begins with the teapot being filled up with water, then hopping over the kitchen counter onto the stove where he gets stuck. The hotter the pot gets, the more he steams and the angrier he gets till he finally explodes.

I tried to apply some of Ed Hooks’ theory to make this story such as starting the scene in the middle, not at the beginning. When we are introduced to the teapot he is already being filled up with water and carrying out a task. His objective is to transport the water to the stove to boil it, his action to get there is by hopping along the counter, until he meets his obstacle and gets stuck on the stove. Now his new objective is getting off the stove before it’s too late, his action to get over this obstacle is to pull and stretch, until finally it’s too late and he overheats and boils failing in his objective.

I knew the hardest part would be to animate the teapot jumping convincingly, so I decided to break it up. I first animated a simple ball bounce to get the timing to how I thought would be right for the teapot. Then once that was done I drew the basic shapes of the teapot on a different layer and keyframed the movements over the bouncing ball. This is where I played with the squash and stretch of the teapot and the rotation. I paid a lot of attention to getting the arcs of its movement right. Since this is a very cartoony story with a living teapot that explodes, I decided I should make the teapot’s movements quite cartoony.

I then used this animation test as a guide for animating the real teapot in the scene. The mistake I made with this animation was animating it on 1s. When I came to animate the rest of the animation I did most of it on 2s, as it looked more cartoony and saved a lot of time. This created a problem when trying to trace this guide later on as some of the nice movements, and squash and stretch was lost, So I had to modify the animation for that.

An important thing I did learn while working on this animation was the importance of flipping drawings like in traditional animation. Working in a traditional animation workflow, I found using the onion skin feature not very reliable especially when animating faces. Sometimes mass would be lost if I drew something working only from the onion skin, so to keep my drawing on model I was constantly smashing the arrow keys to go back and forth through the frames. This is advice I got from a lot of places when watching tutorials on solid drawing and traditional animation.

To make the travelling background, I created a wide canvas in Photoshop and began to illustrate the background on it. I kept it simple as I didn’t want it to be the focus. When I was done I just imported it into Toon Boom and animated it sliding in sync with the teapot’s hops.

I had a lot of trouble when animating the fire and the explosion as I had never done that kind of animation before. So I read chapter 5 of Elemental magic by Joseph Gilland To improve my animation. I learnt that for fires and explosions, a straight ahead animation approach is the best, as these things are unpredictable and can’t easily be planned and inbetweened. So when animating the fire I just let the flames decide which direction they were going in and just followed that. I knew that the fire would always point upwards and bits of flame would break off and fly up, but other than that there wasn’t much structure. The challenging part was making the animation loop, to do that I had to use the onion skin and inbetween. But for the most part, I avoided using the onion skin so that I could make the animation feel natural.