We use cookies on this website. By using this site, you agree that we may store and access cookies on your device. Find out more and set your preferences here.

SVA Students on Taking The Plunge into Production

SVA, the School of Visual Arts, is a premiere art university, well known for their Animation and VFX program.  Over the last year Solid Angle had the pleasure of visiting the school while students were wrapping up final projects, and one film caught our eye for it's endearing story and visuals: Taking The Plunge.  We learned this past summer that not only was it a strong short film, but it was also the winner of the coveted Art Director's Club (ADC) award, a first for a student film.  

We asked filmmakers Thaddaeus Andreades and Elizabeth Ku-Herrero to talk about their process, how they used Arnold and what they learned along the way.  In addition, SVA's Richard Hagen joined to discuss their pipeline and the logistics of rendering so many student projects in a short period of time.

Tell us a bit about your individual backgrounds as students of Animation and Film.  Were you trained in traditional arts before CG?

Thaddaeus - I have a fairly mixed background, starting out wanting to make films and really loving movies. I assumed that I would try to go to school to be a cinematographer or editor or something.  I discovered animation while watching making-of DVDs for movies like Lord Of The Rings, and The Matrix.  I got hooked on the idea that animation could create this amazing visual storytelling that was impossible to experience otherwise.  I downloaded a Maya demo and spent hours just messing around.  I talked to some people who did it as their job, and they all advised a strong background in drawing.  From then on I went to an art high school and took as many anatomy and painting classes that I could.  I finally jumped back into CG for my undergrad at SVA.

Elizabeth - I’ve been drawing as long as I can remember and always believed I was going to go down more of a traditional path. I wasn’t quite aware of what CG was at a younger age, so I took summer classes. In 2008, I was accepted to the Pennsylvania’s Governor’s School of the Arts right before my senior year of high school. I had grown up making live action shorts, Flash animations, teaching myself HTML, but also was a very serious student. I felt stuck between taking the artistic path and living as that romanticized ‘starving artist,’ or make use of my grades and pursuing something like Accounting. CG felt like a perfect marriage of what I was attracted to in a medium; it requires an artistic eye, technical savvy and constant interdisciplinary research.


How much specialization was there with a small team like this?

Elizabeth - Our film was a particularly interesting project to work on because we had declared ourselves as being four directors, though we did have our own focuses. Marie Raoult was in charge of Character Modeling, Rigging, and particles, Nick Manfredi was in charge of Cinematography, Thaddaeus was Animation and Lighting Supe, and I was Creative Director responsible for shading and environment modeling. However, since it was quite an ambitious film, spanning about 70 shots, there was a lot of crossover that needed to happen. Thaddaeus and I decided to split up the lighting and compositing between underwater and above water shots. Marie helped me out with environment modeling and joined the guys during animation. Sometimes it just came down to avoiding bottle-necking. If we needed to push out something in the production in order to move down further in the pipeline, we would redirect work if someone had less responsibilities on their plate. We actually made the joke that we were all 2 legs of an octopus and we needed to learn to lend a tentacle and criss cross arms when needed.

How was the look developed and how did it evolve?  What were some inspirations with the overall look and feel?

Thad - As proper students, we had no idea what we were doing when it came to the look. We took reference from Tangled for lighting. However, for surface detail we looked towards personal heros that we followed online like Pedro Conti and Victor Hugo. I really loved the detail they both achieved when it came to skin. All of their character work had some really rich SSS and what I’ve found appealing is that the shadows are really warm. You never lose your colors to muddy grays. We used the straight-out-of-the-box MtoA plugin for Maya.  I really love how Arnold handles volumetrics which was invaluable for our underwater environments. We also really enjoyed using the Arnold skin shader

Elizabeth - I was extremely lucky at the time that we brought on Eric Cunha, of Aardman Nathan Love, as our Academic Advisor. He taught me the basics of Arnold and I had a range of characters to apply this all to.  Arnold was a breath of fresh air because I could spend more time being an artist instead of messing with render settings. I started with the Standard shader across everything, relying heavily on displacement for the environment. As I became comfortable with shading I dove more into skin shaders. I tried the Standard shader on our Octopus and the shark, but when I tried a skin shader the characters became so much more tangible. The skin shader for the eyes is something Eric had taught me as something that helps sell the characters. I’d say the entire process was done by the seat of my pants- I didn’t have much experience to go off nor enough time to really get more iterations in there. I’m just relieved that my shading chops were enough to compliment my teammate’s work and the story.  Look Dev was primarily done through Maya, Zbrush, Mudbox, Arnold with MtoA and Nuke.


How did you approach production management challenges and keeping the look cohesive?  How were shots assigned?

Elizabeth - Half way through production, before lighting started, I purchased Jeremy Vickery’s Practical Light and Color DVD and it opened my mind to the process of shaping a film with these new principles in mind. As soon as I finished it I worked on a colorscript for the film. I felt that our story was long enough and went through enough transitions that a broad color palette could be used and could enhance the story so much more. Since our previz was completed and animation was under way I was able to paint over a lot of Viewport screen grabs straight from our film. We decided to split the lighting and compositing between Thaddaeus and I, with him being responsible for the underwater shots and I was responsible for the above water shots. My colorscript became our bible. It’s extremely handy to be able to paint something up in 2D and see it work as opposed to working on plenty of iterations across 70 shots. As we were lighting we kept a contact sheet of our shots in progression as well to make sure our film was successfully hitting my color cues.  

What were the toughest lessons learned from the lighting and rendering process?  Where were the big discoveries?

Elizabeth - We definitely learned about the importance of a strong pipeline; proper naming conventions, establishing Look Dev scenes, and the power of automation when your file structure is clean and consistent. We would not have completed the film without Eric's help because instead of working on our film we would have spent half the time just setting up scenes, reapplying shaders, having things break, etc. The night before our film was done Thaddaeus and I wrapped up the last comp at midnight, looked at each other and realized we were done. How are we not in panic mode? How are we about to go to sleep? It was all possible because of a proper pipeline.

Thad - Agreed, the toughest lessons were learned in pipeline.  There were growing pains in the beginning, getting all our shots set up in Shotgun, and figuring out how to push through rig or shader changes to all 70 something shots.  But it became invaluable later on in production. Besides a few exceptions, I only had enough time in the schedule to render each shot twice:  once at half rez, then make the fixes, and again at full rez.  If our pipeline hadn't been working by then, we most certainly would not have hit our deadline.


What kind of render farm capabilities does SVA use?

Richard- The SVA Renderfarm at the time consisted of 25 primary render machines and 150 total - we have racked units which just do rendering and then when everyone is rendering (it was about 55 individual projects, including Plunge) I expand out into the labs which gives me about 150 machines total. Each machine has 32GB of RAM (the racked machines 32-64gb), with E5-2650 processors (32~48 cores each).  Because Arnold scales so well, we take difficult renders and devote an entire node to them while several smaller projects can be rendered simultaneously on the same machine (as long as there's enough RAM to hold all the different jobs).  Average times were an hour or so per frame - we had to try to enforce a time limit of about 20 - 60 mins a frame due to shared resources.  All rendering was done on-site and MtoA was our Arnold platform for our artists with Shotgun being used to manage everyone's projects.

Part of the process for students is learning to use the farm as a finishing tool not as a prototyping tool - i.e. prototype and test locally first, then test some frames on the farm to get a feel for the quality. Once it passes muster, then we grant access for the whole scene.  Other discoveries were the importance of using Arnold with proper materials and lighting setups - the transition to a PBR renderer causes you to rethink a lot of the processes you're used to.