Feature Crew

Andrew Kaufman


R&D Lead Andrew Kaufman discusses his work in advancing Image Engine’s pipeline, how developing tech has made the studio more competitive, and what he finds most fulfilling about his role.

Tell us a little bit about your background. How did you decide to get into visual effects?

I think I decided to steer myself towards CG back in high school. A few music videos had come out around then using CG characters of the band members. I was taking a programming course at the time and asked my teacher how I could do that sort of thing.

I ended up at Northwestern University, outside Chicago, as an engineer studying Computer Science, and taking all the graphics courses they offered… I even co-taught an “Animation Studio” course, which was a bit of an experiment to try and learn Maya, rig some characters, and teach other students to animate a short film, all within a single semester.

Afterwards, I wound up at the University of British Columbia, studying Computer Science again, this time towards an MSc. My research there was focused on muscle simulation, in the Sensorimotor Computation Lab. We were taking a biomechanical approach to muscles, publishing our work not only in the graphics community (Siggraph) but also in the medical community.

What brought you to Image Engine?

As I was graduating from UBC, I started interviewing with VFX and feature animation companies, mainly in LA and Vancouver. I wanted to stay in Vancouver if possible, but the city had very few studios with R&D opportunities at that time. While preparing for an interview with another company, I came across a press release that mentioned Image Engine and looked them up. They didn’t have any R&D positions posted, but I decided to roll the dice and apply as an FX TD. Luckily they looked past my severe lack of show-reel and noticed my research background. They just happened to be ramping up in R&D, in preparation for some movie called District 9. So I signed on the dotted line as a Junior R&D Programmer, and all these years later I’m still here, now the R&D Lead.

What does your role as R&D Lead entail?

My day to day is fairly evenly split between meetings with production supervisors and leads, managing task priorities and reviewing work for the Software team, and getting my hands dirty in code. We strive to be an accessible team here, so all the R&D and Pipeline folks sit in the same space, right alongside production departments. There are a lot of back-and-forth discussions going on on any given day, many of which I’m listening in on and contributing where I can. It’s all about having a global perspective on the work being done within the team at any given moment, and balancing how that affects current productions, always keeping an eye on the future.

How has tech changed Image Engine since you started?

The technology changes have been pretty drastic, to say the least. When I started, Image Engine was transitioning from TV to Film, and we were building a brand new pipeline for District 9. I was the 4th member of R&D, and we didn’t have any Pipeline TDs. We were able to pull off something really special for that show, but it was a growing-pains situation for us, and it only worked because the artists were willing to hang in there while we cobbled things together as fast as we could.

Fast forward eight years, and we’re a much more capable studio. We regularly work on blockbusters with overlapping schedules and incredible scale, in terms of character complexity, environments, fx, etc. The combined R&D and Pipeline team today is 11 people strong, and we have a healthy proportion of Pipeline TDs embedded into the production departments as well.

I think a big part of the reason Image Engine was able to scale up so well is due to the stability of the pipeline we’ve built, with solid foundations like Cortex and Gaffer as the building blocks for all the tools we create. Of course, the another part of that puzzle is the amazing artists and production staff who take our tools, flaws and all, and create something really special with them.

How does technology help mid-sized studios like Image Engine stay competitive?

There’s a big challenge being handed to mid-sized studios right now. You have to figure out how to develop an identity, become known for something, and still stay nimble enough to tackle any work that comes your way.

The major DCC applications like Maya and Houdini, and even renderers like PRMan and Arnold, have become so prevalent throughout the industry that you want to keep your artists within a familiar environment as much as possible. At the same time, there’s a huge amount of open source software available today that just wasn’t around even five years ago.

I think the sweet spot for a mid-sized studio is to take those common applications and those open source toolsets as a base, and build the glue and plugins and added-value tech on top, to enable your artists to work more efficiently, get more iterations on the art without getting bogged down in the semantics. The goal has to be towards artistic flexibility within an otherwise automatable workflow.

What have been some of your favorite shows that you’ve worked on and why?

District 9 has to be on the list, just because it was my first job, my first screen credit, the studio’s first real film project, and we got Oscar nominated! I don’t know how to top that feeling.

Aside from that, Jurassic World was pretty special for me, since the original is the first film where I remember being blown out of my seat in the theater.

And it’s a funny one to mention, but The Twilight Saga: Eclipse was an awesome one for R&D, despite having no interest whatsoever in the plot. We’d been awarded the work a few months before shot production was starting and we had no idea how we were going to pull off a CG wolf. So three programmers and a CG Supe sat in a room for 2-3 months, and we bashed out a proprietary fur system, pretty much from scratch.

What aspect of your work (past or present) are you most proud of?

I’m most proud of our open source tools, Cortex and Gaffer. I’ve spent my entire career being paid to develop open source technology which isn’t all that common. It’s been very fulfilling to share my team’s hard work with the industry, by presenting our technology at DigiPro, FMX, and Siggraph, by making the code publicly available, and by encouraging others to get involved, or hopefully inspiring them in their own work.