Shape's Race Animation Production Blog By Colin Leet


Project Status, September 2019

I only ever ended up completing the first 1 min 50 secs out of 7 minutes I had planned for this animation. By the end of the semester I was working on this full time (Spring 2015) I hit a major burnout on this project, and needed to take a prolonged break from it. In the years since I've only occasionally battled down in wee' little bits the mountain of work this project is; updating the rigs, scenery, completed animation, music, etc. My hope is to still complete it sometime in the next 10 years (by 2024 would be great!).


I ended up changing the name from The Shapes High Race to Shape's Race after the semester ended.


Shapes Race Beginning From April 25, 2015

This is the video from my first public presentation of this project.
It's just the first two minutes of the film.
Download MP4 Link Here


Milestone 6 - April 8th 2015


Goal: Have the film 1/4th animated.


Progress: Done! I have the first minute of the film almost entirely done. The project is finally moving as fast I could have hoped for, with the set and everything primed I've been able to get down about 25 - 35 sec of footage a day know that's a bit fast but almost all of the assembly/thinking work is done, allowing me to just focus on their physical movement. It's so freaking good to finally be back in the flow state again, the constant problem solving I've been doing all semester is fun, but nothing like this. :D


Milestone 5 - Mar 25th 2015


Goal: Make needed changes to story (rework storyboard), start sound design, finalize most non-dynamic scene assets for rendering.


Progress: Done. I was way way too ambitious with my original set of goals in thinking that two weeks would be enough time to finish the character rigs, with them eventually taking six weeks of work in lue of two... (well that's what happens when you switch over to new software suites mid-project). One thing I figured out when I tried moving into production-mode was that I actually never have before attempted to figure out the final tempo of my animations without also having the my music present to inform the cuts and tempo of the final film, so I made it a priority to get that part of this puzzle figured out first.


Music: I ended up choosing a public domain 8 min recording of Bolero. The music will of course be cut down at least by 1/3rd. Sound effects will fall in quickly along the way as a rough cut of the film gets assembled for the next milestone. Plus what is better than a 6 minute crescendo of increasing ridiculousness?


Story: In my failed two day attempt at animating I found that much of the movement I had been developing the character for was actually in slow motion. Rather than just slowing down some shots, an idea came to my mind that the characters shouldn’t necessarily perceive time as moving at the same rate we do. What does time mean to a circle? The film takes on a new quality when viewed at 100th speed (as any animator will know after working on a shot that only preview-renders at 1-3 fps!). At this speed simple actions start to take a new life on their own, becoming more like paintings than moving pictures. As the overall range of motion I want to get out of the characters is now rather limited (as in I can't change it at this point), how, or the rate at which those actions are seen by the viewer seemed to be one of my most powerful tools I have in separating the shape's characters. I'm not going to try to over do it (like slowing down the motion to 1/100th real time) perhaps by just slowing down the speed by 1/2 or 1/4ths normal. I don't know if this is a tact that I'll be keeping with yet, but I think it’s something I'll still try to work into the story. I can always just speed those motions up if they don't work well.


Through some introspection I've come to realize that the main things that I want to show about these characters isn't so much their motivations (they all just want to finish the race), but instead their struggle is with how they control their movement to get their goal. I've been playing with a lot of ideas over the past month as to how I could expand this world, but it was seeming like I was throwing on too much. This world is bizarre enough without me going full blown crazy with it. (Think of these shapes as being part of the hunger games, and/or Olympics, mixed in with a little 1984, and Gladiator... Yup. I needed to be put in check.)


How the characters move was what made we want to animate them in the first place.... I need to keep to keep it simple, leverage my bizarreness to work with the simple plot, rather than against it.


Limitations are more interesting than abilities. Limitations make the drama, abilities solve them. This is a story all about finding out these characters - building expectations, then breaking them. Drama is struggle. Comedy is struggle gone wrong, and this is a film about failed and strange and staggered forms of movement. My biggest limit is that I've got to keep it all respectful of physics, no breaking the magic of the struggle, or the tension in the music.


I've gone over the storyboard a couple more times and pinned down to a much larger extent what should be happening second to second. I know from my past experience that I never finalized these things till it's finally animating time, but this should make informing and making those decisions much easier process along the way. I finally feel like the film has its full trajectory laid out - now I just need to get my nicely primed movement in the environment. >=)


Finalizing textures: I spent around 15 hours reworking almost all of the textures used throughout the scene. Here are a few new picks.





Milestone 3 - Mar 2nd 2015 / Milestone 4



Goals: Stepped animation 1st half / Stepped animation 2nd half


Progress 3: Both Delayed ~ additional design overhauls for the characters kept on breaking the dependency requirements for getting the characters into the final scene / Needing to fix the scene to match the quality of the characters.


Progress 4: Needed changes have been made the scene's geometry such that blocking stage movement should now be possible. However I'll be reworking the story as was my original plan for the next milestone. I've been flying by the seat of my pants with how loose the storyboard is, especially given how hard I've been on myself in terms of the visual quality of this project. The simple truth is that after trying to spend a couple full days animating the scenes without much more detailed storyboards (or at least making additional notes to work from) I ended up not knowing where the shots should be ending/beginning much of the time. Better to throw out a couple days of work and spend some time updating my plan before moving forward. This is a time when it is better to take four steps back so I can actually take five steps forward.


All three character models underwent their 100th redesign ~ although I got them to a good point for Milestone 2 I decided that the models needed to become more similarly styled. I ended up settling on a more flesh like look for all of them (not just Boxy this time). Springy wasn’t working all that well with his half plastic-half metal look. His physical look reminded a couple people I showed him to as that of a coiled-snake, so going off of that I ended up spending several days giving Springy scales via object instancing in a hair particle system. Additionally many of the tricks I learned in rigging turned out to be useful in other areas of the other models, meaning that most everything was redesigned, or tweaked to remove bugs I made along the way. They may not look all that different now, but they're control systems are a lot more stable underneath!


Milestone 2 - Feb 18th 2015



Goal: Make needed fixes on rigs, finish texturing, place characters in beginning and end poses for all shots.


Progress: Characters are 95% there, after redesigning them became a much longer and more involved process than I expected. I've been playing a lot with their animation systems, automating secondary movements throughout the rigs where possible ~ squash/stretch, more fluid bodies, playing with rigidity and fluidity in differing areas of their skin and bodies to simulate different materials, sculpting facial expressions etc. Likewise, with the changing nature of the characters means that their materials (metal, flesh, marble, plastic?) have been in a state of flux. The textures are simple at this point: mostly solid shaders without texture maps, but my goal with these to get a better feel for the characters than a more finished result.


Moody Spring ~ The chin is now moveable in all 3 dimensions via a curl. It also has a 4-point articulating IK Spline which can be blended with the original curl, making it possible to bend it into different emotions, and then manually articulating it into holding on object as well. This appendage was added to the story for several plot points when I finished the storyboard last milestone. Additionally the neck now has a separate IK chain for more controlled neck and head movements.


Boxy -- Since Boxy physically is the most similar to humans in terms of anatomy I've been playing with the idea of giving him a more skin like appearance. The main problem I found with it is that the subsurface scattering has the tendency to hide all of the fine details around his nose.


Much of the work that I wanted to do in getting these characters integrated into the full scene (for today the start positions of each shot) was just to make sure that the rigs would be expressive enough to move emotionally. Much of this just breaks down into a pragmatic problem of trying new rigging styles until I find something that technically works, lots of trial and error. Eventually once I figure out a good way of programming in that movement it ends up largely changing how all of the other shots with those characters feel. They're personality is almost entirely expressed through their physical movement, so changing how their rig moves later on would likewise require redoing any shots I previously animated. It's something you've just gotta get right early on.


I'm trying to keep the rigs as simple as possible in terms of controls. Boxy was animated into this position with only three handles. Several of his walking animations are about 80% fully automated using just one or two sliders for controlling the bulk of his physical movement. (It doesn't automate stuff like his breathing or expression.)


Milestone 1 - Feb 4th 2015


Goal: Completed storyboard and make it into a real-time rough-cut.


Progress: I've completed the storyboard, click this link to view it.


My original intent was to make the storyboard on the rudimentary scenes I currently have in Blender by directly drawing directly into the viewport (leveraging the grease pencils' recent improvements to try to get some rudimentary animation planning in as well). This is as far as I got...


The first 2 seconds looks nice because it took 60% of the time...


The grease pencil tool looks awesome, but the lack of any kind of tweening between frames made working with it impossible for my needs at this point. The tool only currently supports complete layer replacement with its key frames, making it a horrible tool for repeating animations or anything that isn't static in 3D space, as it takes nearly as long as redrawing new frames from scratch than to manually select and drag around the old lines to new location within the frame. Because of this I ended-up finishing the storyboard just on paper, as without having the overall movement and timing of each shot planned out, a full attempted edit of the film at this point of just the still frames wouldn’t have had much meaning.


Moving to Blender

After considering the amount I changed the original story from a year and a half ago I've decided to move the project to Blender, instead of Maya. This means that I'll have to redo all of the extensive rigging work that I've already completed. However, I think it's worth the extra time, especially as some of the story modifications made in this draft require a much different range of movement from my characters than their current rigs could provide. I was fairly unhappy with my original design of them after playing around with them for a couple months. Moving the rest of the scene was much easier than I had expected, a single fbx export successfully transferred all of the rudimentary scenes geometry and texture I had completed. Additionally blender's built in render engine, Cycles, can leverage GPU rendering which makes it much more responsive and a much more enjoyable experience to work with than Mental Ray within Maya. Also, the added resource of Freestyle (an included line rendering add-on in Blender) fits better with the loose style I've been wanting to adopt for this project.