Sophie's Animation Casebook

[sophie-uzm]

Making Of Integrated Practice Casebook

(Creative Collaboration Masterpost)

Planning 1

Planning 2

Modelling

Rigging 1

Rigging 2 (Alt rig)

Overview

Showreel: https://vimeo.com/500466284

[sophie-uzm]

Creative Collaboration: Overview

My part in the project was a big challenge as it was mostly new tasks to me. I know now that I should have looking into modelling for a rig, or found a reliable series of tutorials rather than individual ones, because the biggest problem I had was with rigging because of the nature of the model. I did however learn a lot about rigging which will probably be useful in the future. I sometimes felt that I was limited by time but not too much so, though I felt bad for not getting the rig to the others sooner. The changes to the desert section so it didn’t land were frustrating, and could have been saved by finding a rig that could fold the wings in online but it would likely have felt wrong to use someone else's work in our project. 

The restriction of staying at home was mostly about motivation for me. the remote desktop caused some trouble but mostly I found myself easily distracted by nothing, or just not putting the effort or time in because i felt too comfortable at home. Luckily this being a group project my main motivation came from the responsibility and anxiety to provide what I needed to to the project. We had a couple disputes one of which required a cool off period of a few days, and made communicating awkward at times, we were well into working so it didn’t hinder us much.

The project as a whole I am extremely happy with and the animation of the owl turned out great despite the rigs restriction, and I am extremely glad with how the model looks lit and rendered out in most of the scenes. I think the plan to focus on visuals over story really worked, and shows in the final animation.

[sophie-uzm]

Creative Collaboration: Alternative Rig

the standing rig was made because the rig couldn’t fold it’s wings in the way it needed to. This was the rig/model could be swapped between shots, though it would be hard to make flow well that way. This worked decently in most scenes that the rig was used in, the hardest being the desert scene where it was supposed to land and set off in the same shot. The part had to be changed from the storyboard and it was disappointing to remove such a nice shot. I had improved the head and leg movements in this rig because they would be more likely to be moved when standing and the wing section of the skeleton was removed. This didn’t come in to play much however as most of these shots were added later and didn’t last long so weren't given as much attention. None the less it was more experience and practice with the rig. Changing the model was also interesting and harder than expected.

[sophie-uzm]

Creative Collaboration: Rigging

I had to entirely follow tutorials for rigging, as I had seen rigs and skeletons before but not made any. I managed to find a couple for rigging birds that I followed. the skeleton took a few changes presumable related to the difference between the model in the videos and my own. Eventually I had a structure I liked, which included separating the head joint from the parent joint and rest of the skeleton to give it more flexibility after testing it. A lot of skin binding settings I was recommended to try or wanted to test wouldn’t work because I made the model not knowing the right way to do so for rigging a character. I made some changes and found one that did the job best, but I’d aim to learn from that issue in the future.

Weight mapping was made hard for the same reasons, as sometimes parts that were weighted wrong would be hidden until I moved the joint. Part of fixing this meant adding a lot more edge loops to the legs because otherwise they moved stiffly and the base often clipped through, other parts just meant moving body parts to the furthest point that might be needed and finding the problems to fix. Some of this however didn’t totally recover and the rig was pretty limiting.

To try and make up for the problems I also added NURBs that would hopefully make the animating much easier and smoother. A lot of these were attached to IK joints, which I think was most beneficial on the wings which would be moving the most and since it would be gliding more than flapping a lot, its restrictions shouldn’t be too obvious or frequent. The test animation Brandon did with it looked good so I’m glad it worked out. (file having errors uploading to post but is in the showreel)

I was frustrated with the rigging stage and guilty for making the others use it, but I’m also satisfied with it considering an unsuited model and lacking practice. 

[sophie-uzm]

Creative Collaboration: Modelling

I do not have a lot of experience modelling other than basics and the tools, so I found a simple owl model as a base. It did not have any texture or detail so I could add to it rather than totally using someone else's design.

most of the feathers are very simple flat plains I shaped, and for some I applied smoothing like on the face so the ‘feathers’ were nicer and easier to look at. I planned to for the rest too like wings and back, but decided after a chat with teammates to leave them be as it may match the low poly style in the scenery and speed up rendering even if only slightly. Certain parts were pretty hard to figure out like the claws where I ended up using cones, cylinders and then surfaces that cover the front and back of each, a situation where building up from the original base rathe than unique shapes would really have helped. On the other hand feathers using individual shapes was likely faster for that aspect.

The modelling took a long time but I’m glad I spent time on it as visuals was the focus of the project, and I’m very happy with how it turned out. I didn’t make the model with rigging in mind which was difficult in the long run but as a model alone it worked great and the design fit into the scenery made by the other group members.

[sophie-uzm]

Creative Collaboration: Planning

We knew the modelling of the islands would take a while which is why instead of giving one or two the role we split it up. Since there were 4 islands the others all took one task each and I was assigned the owl. We also used Trello for the first time which was easy to use and likely helped us keep up to date and organised, though we used it much less later in the project.

I tried to make the owl’s design full of simple shapes so I could translate it to the model easily, and similarly limited the colours. The wing pattern was hard to figure out but I at least had an idea of where to start, and having the structure of the face was definitely helpful. I might have been able to make a better and more reliable reference if I had spent longer on it, but went onto the modelling fast so I had longer on it.

[sophie-uzm]

Creative Collaboration: Brainstorming and Experiments

At the start of our planning we used Microsoft whiteboard, which despite it getting very laggy the more we did, really helped with communicating ideas. We had a board of images and sketches that related to our different plans, some of which had potential but we didn’t choose one immediately.

At the planning stage, each of us also did an experiment with Maya or Blender to give us an idea of what we were able to achieve. I did some tests with the Maya Bifrost system, as a few of our ideas involved water. I leant it using a tutorial on YouTube and found it reasonably simple to learn but hard to make more complicated animations with. I didn’t end up working with this for the animation but Bifrost water was used and I think it works reasonably well.

Frankie's low poly tree models which were his experiment are what inspired the final idea. We had an idea of floating islands from a brainstorming session and decided that we could focus on visuals by making appealing low poly islands. For story we knew we needed something to follow between them and decided on an owl inspired by some owl designs I found online that looked like they were made of paper that fit nicely. Because of this I also offered to focus on designing the owl once we started planning.

[sophie-uzm]

Performance and Character Master List

Sword in the Stone

Scared by a Spider

Flower Sack Rig

Broken Computer

Overview/Reflection

Showreel (vimeo)

[sophie-uzm]

Performance and Character Overview

I think the biggest thing I learnt from these exercises was how to use sync sketch and how useful it was. I doubt I am using it to it’s best yet but it was rather simple to pick up and sped up the process a lot. The flower sack and sword animations each strayed from their references but greatly benefited from having them for timing as basic blocking/posing. Meanwhile the spider animation was almost an exact parallel to the reference, simply with tweaks and duplicated steps I added once it was mostly done. This one ended up being the fastest to complete and also the one I am most proud of.

This also relates to that I like the feeling of realism. While most had some touches of theatricality at least, the two that were simulations rather than representations felt and ended as the nicest. This may be because CG animation is a little harder to manipulate and freely interpret than 2D, but I definitely felt the different style for the sword in the stone animation was a little harder and more uncomfortable to try and animate. The inanimate object animating on the other hand, while unrealistic, still felt like a close reflection of the reference and the interpreting what it’s movements and behaviour would look like was rather interesting, though a bit harder than the bucket kid rig.

Time management also come into play a lot for this module. I put work off a lot in favour of another project because of a sense of guilt to do well for the group i was working with, but I could have divided my time better, as cramming it into the break would likely have depleted the quality, especially as I started to wear out from too much work on it consecutively. This also meant hitting problems with the remote desktop was much more dangerous of a problem than it should have been. I found that sometimes the one drive wouldn’t sync from it, at one point loosing a few hours of work, and opening different apps would open new desktops rathe than on the same one. In the last few days it also started freezing for up to ten minutes at a time. Most of this seems to be trackable back to using it for too long in once session, but not realising this cost me time and probably motivation and work quality as a consequence.

Despite problems and personal regrets, I feel I have progressed with certain aspects of the animation process, especially with the details of the maya graph editor and using references.

[sophie-uzm]

Performance and Character: Broken Computer

Inspiration

This ones inspiration came from videos I’d seen poking fun at a task from the game Among Us that people often find frustrating. Like from 6 seconds in this video or 15 second into this one. I like the joke of a character getting increasingly frustrated with a simple thing not working. I also decided the character could be not particularly happy to be doing the task at all, probably at a 9 to 5 job they're sick of, meaning they have a short temper with it. And as a way to end the short story of the bit he’d realise it wasn’t plugged in at all.

Planning

I recorded this one in front of my real laptop at  my desk to make it accurate even if it cause bad lighting and I had to avoid my bed which wouldn’t be there in the imaginary office setting. I marked down a bit less for this reference but still did a few keyframe sketches on sync sketch. Due to facing the camera sometimes it was hard to tell the exact angle things were at, especially my arms. By this point it had become rather routine to use the website and I didn’t really do anything differently.

Syncsketch annotations

Animation

When working on this animation I was getting close to deadline so it was somewhat rushed in places, despite this I am satisfied with the result. the first and possibly biggest issue I ran into was the step up to the chair. I found that because I couldn’t see my feet in the reference I wasn’t sure where the feet went, which was possibly to infer but upon running through it it would clip through the arm of the chair.

I created the props as simply as possible but as chairs are all different and specific dimensions I couldn’t find the right middle ground. If the arms/whole chair was too far back bucket kids arms didn’t sit on the arms right and if I moved it forwards the leg and hand would clip through at the start. I went for the former issue and just used the chair boxes as a guide. Because of this the hands seem to grab onto nothing to pull the chair into the desk but in reality it is meant to be the arms of the chair. I tried to get this across the best I could and it is definitely something I’d return to and try to fix with more time.

^ At this point and the keyframes around it it should be holding the arms of the chair but they’re moved back for when it steps up to it

The animation beyond this point went reasonably well. I wondered if I should have used IK controls, which would probably have helped keep the inactive right arm in place when the body and left arm moved, but for the act of hitting the keyboard it probably allowed me to be more detailed and accurate in the movements. Bucket kids simplicity has appeals to me quite a lot although the nature of the model makes interactions with solid surfaces slightly odd.

This last animation would have definitely been improved by spending more time on it which is my fault for not starting them all sooner despite the other modules, but also was made worse by issues with the remote desktop system. Considering the time I am okay with this ones result, but compared to the other animation with this rig, the spider one, I know I could do much better, such as the number of times the torso is too still or moves robotically.

[sophie-uzm]

Performance and Character: Flower Sack

Inspiration

I wanted to use the flower sack rig that was recommended to us and tried to think of an idea to match. I thought about animals and also about small children as the small size and clumsy movements of a toddler would match better than that of a grown person. that is why I made the character personality that of a young child as well, specifically a curious and playful one, creating the idea of it climbing onto a box just for the fun/challenge of it.

Planning

For the reference I hoped to find one of children in a play area or similar online. I got some related ones but decided to record something myself to help more. It is split into two parts just because of space in the room to film.

Part 1

Part 2

trying to ensure I acted in a way I could copy to the animation was difficult and toon some trial and error. I hopped around as I knew it would be hard to replicate a walk with the sack. When I annotated I used my torso as the sack and drew lines to decide when it should bend for things like jumps, plus some lines to help guess where the corner tassels would move.

Animation

Animating was confusing at first because it was a new rig but was quite interesting and doable once I got used to it. The lack of limbs possibly made it much easier. I used a singe jump from my reference to start, which ended up in a loop as I hadn’t timed the movement yet which was good to adapt to the rig.

I often found the real timing for a person jumping looked too fast possibly because the flower sack is smaller and it can’t jump as far so a lot of it was slowed down. The bottom tassels don’t play much into the movement, mostly working as feet, and the top ones alternate between representing what my arms do in the reference and just being a secondary motion falling behind.

Interpreting movement’s into an inanimate object was really interesting to do, I often had to change parts. the largest change was from pulling my legs onto the table where I made it arch its back to pull the ‘feet’ up rather than arch it in because that action required my arms. It took times to change that I possibly could’ve avoided if I had considered it better before recording reference. the rest of the reference however came in really handy despite the use of an object. The other helpful aspect was being bale to make it emote with just it’s actions. I felt that creating a young and curious personality was quite easy to convey using the tassels and timing/exaggerating movements.

The rig was also a nice change and felt smooth to use. I ran into a glitch where the top tassels twitch in for a moment despite there being absolutely no keys or curves appearing on the timeline/graph editor, during the pause before it jumps onto the box. Trying to remove it got difficult so I left it in as a sort of fidget of anticipation.

[sophie-uzm]

Performance and Character: Spider

Inspiration

The inspiration for this animation came from early on in the module. We looked at a video where someone is distracted by has phone while testing soemthing so when a bucket falls loudly he is surprised and reacts. To vary the animations I wanted to use different emotions and knew fear would be a good one to use. One of the jumpiest reactions I knew well was to spiders, and I used the idea of reaction to a spider because then the video reference would hopefully be realistic.

Planning

The main reason I needed the reference was to catch the small jitters and reflexes people would do when afraid of something. I made several recordings but chose the one where my twitches were most notable. I think this ref is the reason this animation is the one I’m most happy with and I hope to ensure I have better references for other future animations. Like with the other three i used sync sketch to pick out key poses.

Synsketch anotations

Animating

The hardest part was that the reference was slightly short and I spent time offscreen because there wasn't much space to film in. I duplicated my sidesteps and steps when running away which is probably the biggest flaw in the animation as it is repetitive and a little robotic.

There was some improvisation involved when translating it to an animated rig but has the least. It is mostly a copy of my actions because I wanted to use realism, thought it had some dramatization in.

Aimed for simulation more than representation, the opposite of the sword in the stone animation. this was made a little hard by bucket kid as a rig being quite simple so details in the reference sometimes got missed out, other than that this was the easiest to do thanks to sticking to the reference. I mostly worked by blocking out key poses then looking for inbetweens. For example near the end of the reference I flick my arm back towards the ‘spider’ for a moment as I'm turning away, This looked almost like a glitch when playing it back with the other keyframes so instead I ignored it only added it at the end, making it rather different from the original but really improving it.

^ The matched up key frames from reference and animation with the difference in the hand reflex

I realised while doing this animation that it would have been made easier if I used the rigs master control for all the times it changed direction. I left it alone to save the time it’d take to make sure it rotated and moved in time with the rest of the movements, but no using it seemed to mess up the feet and arm controls as the XY and Z rotations weren’t on the right axis most of the time, causing weird flailing arms and wonky feet/hands. I managed to learn from this for the flower sack animation. Other than a couple minor things I’m generally proud and happy with how this animation ended up.

[sophie-uzm]

Performance and Character: Sword in the Stone

Inspiration

The idea from this one initially came from wanting to test an interpretation animation style, meaning it would be cartoon-like and exaggerated. I had not tried this style before and also knew if I was to use Maya there was only so many ways to exaggerate with a rig as I don’t know if it’s possible to add smear frames and couldn’t manipulate a rig as freely as is possible with a drawing.

^ I planned to use the right end of the spectrum as much as I could

The first idea to come to mind was the exercises from the animation 102 module, as the push animation had been my most cartoon style and a pull might allow me to stretch the rig, ensuring it was not realistic and allowing to make it more over the top. This is also my reason for using the Stewert rig for the first time.

Planning

to make a reference for pulling a sword I had to improvise what to grab onto because I don't have good enough acting to pretend to pull on air. I used the overhanging edge of a unit in the kitchen which had the cost of my hands not being positioned right but should get the rest across. This was pretty easy to annotate in sync sketch as yanking it had clear steps with anticipation and extremes. The fall back at the end was harder to decipher the keyframes and breakdowns of so I possibly should have recorded more versions of found other fall videos from online to help.

Syncsketch annotations

During this references annotations I wrote down the frame numbers of frames I wanted to remember, keys or not. This was useful to jump between them with the sketch jumping buttons but was frustrating as the video was in 30fps so I was putting the frame numbers in a calculator anyway (which I also did on the other ones)  Some of the frames are timed a little different anyway in the animation.

Animation

I downloaded a new rig for this animation, so I did have to figure out how it worked. Bucket kid had a stretch function on the torso but the stretch mechanics of the Stewart rig seemed easier for this particular animation. Learning the rig was rather difficult but it is much more detailed than BK, making it a lot more time consuming to ensure it all looked correct but likely helped the quality of the end result. because of this I wish I had left more time to spare on it, letting me test and learn the rig more, and use its controls to refine the animation. 

The animation process for the exaggerated style was different than usual as i wasn’t considering real life much. I used the reference for timing and some poses but most other poses I improvised to make it more over the top. I swapped the arms to inverse kinematics upon grapping the sword which definitely made the pulling action much faster as i could stretch the arms as much as needed by not moving the wrists. The thing I regret about this was not using this the whole time. The steps up to the sword still use FK which generally would be better but the swap between the two didn’t work smoothly and was hard to get right.

The steps at the start as a whole is definitely the worst part. I tried to really emphasise the confidence of the character on a big march at the start but trying to emphasise the steps seemed to make it seem too unrealistic and no longer looked natural. Perhaps if I was doing the interpretation style again I would carefully choose what to and what not to exaggerate, though it might take some trial and error to learn. I am not sure if I liked doing this style, particularly in 3D.

^ Extreme frame, which has the arms stretched

The position of the hands on the sword felt particularly off/wrong but would be hard to move partially due to the models hand design and also moving the wrist as it pulls would take lots of adjusting the hand grip carefully. Its something I would maybe aim to fix other than the walk if I went back to it. However for working with a new rig and in an unusual style for me I’m satisfied with the result, while being aware of quite a few problems.

[sophie-uzm]

Intro to Making Of: Masterlist

Note: I am missing evidence for a few entries due to a sketchbook i used being at my flat that I can’t get due to quarantine

Pixelation 

Backgrounds:

Background art

Moving Background

Interior

Character:

Concept art

Turn around / colour ways

Maquette 

Review

[sophie-uzm]

Making Of: Overview

I overall enjoyed this module and feel that I have gained a lot from it. The character design was my favourite section as I’ve always enjoyed the idea of designing characters but have never done so with a proper process e.g turn around and colour ways. I’d also never made a maquette which I hadn’t really thought about but actually enjoyed it quite a lot and was proud of the result.

The background was similar in that I’d never fully done it before, but I generally find drawing backgrounds harder so I struggled more. The lessons on perspective drawing were definitely useful for this, but the brief I chose didn’t allow me to apply this much. I also couldn’t get the moving background track to work well which was frustrating. The Pixrelation task was also difficult but more because it was a big group that cause some stress. Despite these issues I still found the tasks interesting and informative. Especially the backgrounds as I would like to keep working at this area to improve.

The final task i had to change my topic from VR to Pixel art but I am actually very happy with my outcome for this. It turned out to be really fun to work on the animations and while it was surprisingly hard to research, as a lot of sources looked more from the gaming side of things, it was interesting to learn about. The power-point itself was nice to work on and I think it looks pretty presentable.

I had lots of different feelings and experiences through the different parts of the module, but I definitely learnt some new things and got an Idea of things i like/ dislike and can do well/ struggle with. 

[sophie-uzm]

Making of: Pixelation

The Pixelation group project was the first thing we did for the module. It was also the most I’ve worked in a group before. The planning originally went decently well. We spent time considering what we could do that the pixelation process allowed and decided teleportation would be fun to include since the person could just walk away from camera between frames.

We made the story around this idea, having people chasing someone with a teleporting cube. This was planned to be the tesseract from marvel but due to some difficulties trying to make the prop, and us deciding to add a battery charge to it to help the story, it moved away from this idea. This did cause some conflict of ideas in the team but this was rather quickly resolved.

We also did some experimenting with the chase scene, as we knew running frame by frame would be difficult. We added someone sliding on the floor as a joke in the short since the method allowed for that, and figured out how we felt best recording the group running since it was pretty difficult. Filming itself took a lot longer than expected and we did so over two days. Some parts needed repeating after mistake or changes to story and we hand’t fully planned the locations. Other than this it went okay though.

The last stage, editing, a few members of the group worked on together and then compiled. This was also pretty difficult in places because of issues and jitters we didn’t notice when filming, and trying to keep consistency between the filming on different days.

being a rather large group meant we ran into a few disagreements but none of these damaged the process too much and the rest of the time we worked fine together so it went pretty smoothly on most days. As for the final product, the animation itself was more smooth in some scenes than others often depending on what actions people were doing on screen. The story seemed to come across pretty successfully though. This project had good and bad points but generally went well for one of the first tasks of the year. it was also nice to try out a process I hadn’t considered before for the first time and see how it works (and the difficulty of it)

[sophie-uzm]

102 Master List

Posing

Pose to Pose

Walk

Stand/Sit

Sneak

Push