Thursday, 2 April 2015

Computer Animation or: How I Learnt to Stop Worrying and Love Pandora's Box

It's been some time since I updated my blog with any detailed content, so here's something to help break that silence. Over the past year or so, my freelance work has changed a fair bit and I'm starting to work on things more attuned to my style. I've also been working on non-CG projects, which have been helpful to get my head out of the game and reassess the best ways to work as an independent artist.

With all of this in mind, I'd like to start talking more about the things I'm learning and the things I care about within my work life. So, to start, I'm going to discuss a recent project and how thoughtful art direction / methodology is making my life way more manageable. In addition to how this is enabling me to find a new way of working in the computer animation field.

In January of this year, I completed some work for an independent film which has significantly helped me rethink methodology and art direction. Until now, I just kind of made CG assets to look the way I expected them. I tried to art direct to an extent, but I guess you could say I was still finding my feet. So, for a long time, I never really felt like the work I made was expressive of me. It's been a weakness for a long time, but I think it's changing. 

The big shakeup during this project was time. The project deadline was really short. Like really, really short, with only a few weeks to complete a full sequence. This meant it would be impossible to make the film using my traditional pipeline. There wasn't time to render, or make smooth, animated characters. There wasn't time to UV and texture paint assets. There was only time to be as efficient as possible. Economy was key.

The director expressed a real liking for some of my paintings, so instantly there was an aesthetic to my drawing that needed to be captured within CG. This was great news, because I suddenly realised my drawings are expressive of me, whilst my CG work so far, tends to feel distant to my intent. So, I sat down and tried to figure out how it may be possible. The first thing I did was look at other artists who seem to be in control of there work. So, I reread David OReilly's paper Basic Animation Aesthetics for guidance.

The importance of animation aesthetics is such a subtle yet vitally important one. It might seem superficial to discuss these things, especially because cinema is so much more to do with content and story than a pure aesthetic experience, but nonetheless the visual nature of animation calls for debate on the subject. There is a continuous raft of animation, both commercial and independent, which looks the same, and I donʼt believe it has to be so. The more we think about the subject the more playful and interesting computer animation becomes, the medium feels to me like a recently opened Pandora's box which is still being examined, understood and tamed.
David OReilly

Preproduction Character Sketches

OReilly's outlook, that the software is a Pandora's box waiting to be explored, is hugely relieving. I used to be a man who clung to rules. I worked the way I was told and followed the rules at every turn; and I was never happy. But this has never been the case with drawing or painting, those mediums always felt expressive and exploratory. The complexities of CG require rules, but for a long time, I'd been a slave to those rules. So, with this project, I decided to just see where things went (I know I've said this before). I started with a quick character sketch based on the director's description and preferences to help define a solid art direction. Instantly, the director was on-board, so it was time to open Pandora's box and start making things. This usually is where the problems start...
One of the main problems with 3d animation is that it takes so long to learn and then to use, from constructing a world to rendering it. There are many knock on effects of this, mainly it prevents people from attempting to use it and employ it artistically, the process is very discouraging for the individual to go ahead and make their film. Simple changes can take hours to do, and very often the process is so rigid it doesn't allow any changes at all.
David OReilly

These problems are hugely stressful and discouraging as an independent animator. The time waiting for renders to tick by takes its toll and it is rare for Maya to feel like a creative medium. So, the time constraint of this project actually led me to a new way of working which is incredibly exciting. Rather than rendering using conventional techniques, I used hardware rendering, which captures the scene directly as it appears on my computer. You move the character or lights and they react instantly. What you see on screen is what you get. Obviously, you sacrifice a lot of visual fidelity but the payoff is refreshing, flexibility. The addition restraints actually lead you to be more creative. It seems like a really obvious workflow now, but no one had ever mentioned it to me as a viable way of film making.

This new approach of instant feedback film making was incredible (although I feel dumb for discovering it so late.) The one caveat however, is aesthetic understanding. A few years ago, I would have struggled to make a scene look good in hardware rendering. It's a bit like watching Blue Peter as a kid. The presenter would make something out of egg boxes and paper-mache and it looked great, but when you gave it a go, it looked like shit. Therefore, you need a certain level of visual understanding in order to make things work.

So, I got to work making some really simple characters and objects. Because the aesthetics were cruder, I had a lot more room for imperfection. I didn't have to spend 20 hours refining the vertices or adjusting edges. Once I had a few assets in software, I started testing out how to implement my art direction. Rather than using UV layouts and textures, I used vertex colouring to paint each face of a model. Again, it's a simple approach but instantly worked. These early tests were inspired by the amazing Eran Hall. Bold, simple colouring to make characters feel almost 2d.

This approach then transitions across to the production of the environment. The environment of the film was an abstracted void space, so it was open to interpretation. Hardware rendering made this super easy. I could model things, vertex paint them, light them etc and instantly see the results. It forced me to try out new techniques, bodging ideas together to make the visuals I wanted. For example, rather than using a glow (as seen on the table) I just vertex painted some geometry with transparency. A lot of the projects I work on involve hyper-real CG, which means spending crazy amounts of time test rendering. This time, things came together pretty quickly. It all felt so much more natural and uncluttered.

And then it was time to rig...I stuck to my rule of economy. I kept the rig simple, partly because of time, but also because I didn't want the headache associated with slow, complex controls. The low poly nature of the character allowed for really nice deformations as well as simple animation using blendshapes. I had animated on 2's for a previous project and wanted to use the same technique. It felt natural to the aesthetic quality of film and also is becoming a personal preference. 

I can't show the film online, and unfortunately, some unforeseen circumstances affected the project and the deadline became even shorter, so the final film was more a complex previs than a fully animated sequence. However, as an exercise, this was huge. The restraints and pressures that came along with this project forced me to analyse my work in the best way possible. I'm guilty of just making things because I think it might please someone or because I assume that's how things are made, so this allowed me to almost step outside of my own head and reassess my whole body of work.

Wireframe Character and Rig
Lately, I've been working on a lot of non-CG projects which have led me to think about my future work. I've still not made my own personal film since graduation (and I'm not necessarily in a rush to do so) but when I do, this will be the route I explore. David OReilly's idea of software as a Pandora's box reawakens excitement for the medium. In addition, the realisation that my drawings are most expressive of me gives me a new route to explore. Lots of this stuff all sounds so obvious, and that's precisely why I wanted to discuss it here. Most of my life has been spent just doing what was expected. Perhaps this came from education or just my own mindset. It is only now that I start to question it all. Perhaps it is the reintroduction of traditional artwork that brings back this sense of exploration. But whatever it is, I'm on board and I'm keen to see where it all goes.


  1. This is really cool to see and a definite 'thinking outside the box' story, the character looks charming and it's always awesome to discover new ways of working, thanks for sharing :D

  2. Thanks Charlie :) Will try and show some more from other projects soon.

  3. Hey Jordan, this is a very interesting and inspiring post and very nice hearing about your experience, a sort of "artistic enlightenment" :D Excited to see more of your engaging characters :)

  4. Really inspiring! It's amazing to see that you are finding your way in all this madness and that it's making you feel satisfied : )
    I hope this road continues on being successful!

  5. Thank you Anass and thank you Sam! Glad it proved interesting.

  6. Lovely! Have you ever released a tutorial or assets on creating a rig like this? I love Eran Hilleli's work too and would really like to see under the hood. Is it rigged with advanced skeleton?

    Thank you!

    1. Thanks! :) Unfortunately I haven't, however, I've just started work on a new project with a similar art direction. So, when I begin production on characters I will try and document some of the production. In truth, I've found that really simple rigs work best for low poly characters like this. They have a great deal of flexibility. I'd certainly recommend checking out David Oreilly's character rigs as a reference.