Adapted from the 1993 children’s novel by Lois Lowry, The Giver is a coming-of-age story set in a futuristic utopian society known as the Community. Having turned its back on war and suffering, the Community idealises the concept of “Sameness”, a state in which emotion and memory – as well as differences in physical appearance – are suppressed. Elected to the honoured position of “Receiver of Memory”, a boy called Jonas has his mind opened to a past that has until now remained a mystery, thus triggering a journey into enlightenment and, perhaps, escape.
Method Studios was responsible for the majority of the visual effects work on The Giver, delivering around 300 shots. Their contribution included creating the main Community environment and a sequence in which Jonas is pursued by a remote drone, as well as a number of one-off shots.
Method also shared assets with Mr X. Inc. in New York for additional shots as the project progressed. Method’s VFX supervisor Mark Breakspear was on set for most of the production shoot, supporting overall VFX supervisor Robert Grasmere and VFX producer Pablo Molles. The Giver was directed by Philip Noyce.
Watch Method Studios’ VFX breakdown reel from The Giver:
Mark Breakspear Talks The Giver
Thanks for sharing your experiences on The Giver, Mark. Can you begin with how you first got involved with the show?
Method Studios has a long and excellent track record on environment-type visual effects. And we had previously worked with production visual effects supervisor Robert Grasmere on Salt, which was also directed by Philip Noyce.
Robert reached out to Method early on. We talked about the kind of work that would be required and the general approaches that would make most sense, given what we knew about the story at that time. It was early days and we didn’t yet have a production designer – Ed Verreaux joined soon after – but we did have some very early concept drawings for the Community. We focused on how we would create that world, based on the book and how Philip wanted to take the look.
The key concept we had to deal with was that of “Sameness”: the idea that no one person, and no one object, should stand out from the others. People wore similar clothes; homes all looked the same; even the geography of their world was mirrored along a central spine.
We met several times in LA to talk through the major VFX shots: how best to shoot them and, more importantly, how to afford them. VFX producer Pablo Molles joined the show and we began to plan the shoot, deciding how Robert and I would oversee the VFX shots, and how the shots would be turned over so as to be ready for any previews and trailers that might come up.
How did you go about designing the futuristic environment of the Community?
At the beginning of the project, the initial brief asked for the Community to be located on a 1km-wide mesa surrounded by impenetrable clouds. Ninety percent of the movie was to take place in this location, so its believability was essential. The production would shoot in many different South African locations, so the digital environment needed not only to be photographically convincing, but also able to work alongside multiple locations and blend in without trace.
The art department provided us with an initial layout of the Community, consisting of all the key landmarks and natural elements: trees, grassland, lakes and so on. While shooting was underway, the team built a basic previz model. On viewing this, the client quickly decided that a 1km diameter mesa was not large enough. We tried several scales, going up to 20km, and finally settling on 10km.
What visual references did you use?
Philip sent us reference of “bits” of places – parks, homes, forests, fields, roads, views – and we used them to develop a bigger picture of what the Community would look like. It was the superlative of many locations all rolled into one, but always based on the “Sameness” concept.
That worked well, but often what makes one corner of a park look really great is its relationship to another section of the park that doesn’t look as good. The yin and yang of a place. When you strip away the yang, the yin just doesn’t have the same pizazz. CG Supervisor Anthony Zwartouw and Matte Painting Lead Jeremy Hoey built a pretty amazing system not only for building the levels of detail in the Community, but also allowing for changes to be made without having to go back to the drawing board.
Also, the Community is artificially flat. This was something we fought over because perfect flatness didn’t look logical at this huge scale. Production was confident that this would play to the “Sameness” feel and, in the end, I believe it worked. But it required a huge amount of detail.
How did you go about building the assets?
The Community environment needed to be full CG, as it would be featured in over 100 shots, ranging from wide shots of the entire mesa to helicopter shots skimming over buildings. The design progressed over several months from a natural, random environment – scattered wooded areas and grasslands with buildings mingled between – to a highly structured, geometric landscape that was almost completely symmetrical.
This posed an interesting challenge, because one of the methods we normally use to create realism is to introduce natural randomness into our work. But the gardens of the Community were all very ordered, with trees gridded and equidistant. So we developed scales of randomness, giving the illusion of rigid symmetry from a distance, but up close revealing the uniqueness of the individual trees and pathways. This allowed the fully practical locations – which were anything but symmetrical – to blend in with our digital Community.
How big was the final Community model?
Given the range of shots we had to create, we needed a way of populating the environment with high-resolution assets on a huge scale. This amounted to half a trillion polygons overall. The assets team, led by Kyeyong Peck, created 40 types of trees (around a million polygons each), plants, street furniture, dozens of buildings, people and the landscape itself. The latter comprised cliffs, urban areas around the main structure called the Odeon, manicured parks and the main living areas.
Did that give you a rendering headache?
V-Ray is our main render engine at Method, so we used a plugin called V-Ray-Scatter, and Houdini, to meet our specific layout requirements. Scatter enabled us to multiply an asset thousands of times without hitting the usual memory limit. Scatter also had tools to vary the assets in scale, rotation and colour, so to not create glaring repetition. The environment lighting team, led by Jon Reynolds, took on a lot of this work on top of their usual lighting duties.
For layout, we developed a system in which our DMP department created a graphical map to define asset placement on the environment. This was then fed into Houdini for processing, and placement geometry would be spat out into Maya, where Scatter would apply the asset geometry at render time. The cloud layer was simulated in Houdini by the FX team, led by Ian Farnsworth. It was set up to encircle the entire mesa. This enabled us to put the camera pretty much where we wanted.
It sounds like a monumental task.
The Community was huge, I mean huge! We had millions and millions of trees in that thing – all different and all modelled down to leaf detail – thousands of lamp posts, benches, people, houses, bikes, motorbikes … The details mattered. It wasn’t an over-built place, but our wide shots showed every little detail and, because we didn’t know which areas would be featured early enough, we had to build everything.
In the end, this allowed Philip to create new shots that the story needed, like close fly-overs of the houses with people going about their daily lives below. It was a daunting build, but Anthony’s CG team pulled it off and then some.
Our biggest shots of the Community weren’t even known about until a month before the deadline. It wasn’t bad planning – it’s just that the cut was changing subtly. There was a need to tell part of the story in more detail, and a picture is worth a thousand words. In just one month, we turned around five extra huge Community shots, all full CG with people, vehicles, clouds … everything.
Tell me about the drone chase that happens near the end of the film. Was the drone fully CG, or did you work with practical elements?
A drone was built on set, and initially it was decided that the practical model would just need “enhancing” and the addition of side wings – something both Robert and I had one perpetually raised eyebrow about throughout the shoot! It’s not that it was a bad design. It just restricted how we shot things because it was so big and heavy; you could only shoot it upside down for the most part and never move it around.
So Robert and I hatched a plan to completely replace it in post and, as we went along, started to shoot more and more shots entirely on greenscreen. The reason it wasn’t planned that way from the start was that the drone initially had a far smaller part to play. As things developed – and long after the budget to build it practically had been assigned and spent – the story grew, and the drone needed to do more than was originally thought. If nothing else, in some shots, it was great to have the correct shadows fall on the practical object, and it was a great tracking object for matchmove. But, in retrospect, it should have been just a simple box on a greenscreen.
Can you describe how the drone chase plays out on screen?
During the sequence, Jonas (Brenton Thwaites) is escaping from the Community, pursued by a drone being piloted remotely by his former friend Asher (Cameron Monaghan). The drone catches up with Jonas, and pulls him up in to the air ready to drop him from a great height.
Many of the shots done with Jonas and the practical drone were quite static, despite being “in the air”. We spent a lot of time replacing the practical drone with the fully CG one, adding in new movements such as pitch and yaw, air compression from the suction devices that lift Jonas, and engine rotation. The end result was to take a sequence that lacked visual energy and give it a huge boost.
We spent a couple of days shooting motion capture at The Capture Lab in Vancouver to add a CG version of Jonas into some shots. Our digital Jonas looked great – the digital double team, led by Chris Norpchen, did a fantastic job of getting all the hair and cloth detail in and working well. To top it all off, it actually looked like the actor. You assume the scanning and texture maps would take care of this, but it doesn’t always work out that way. When you look at the shots and wonder which ones are digital Jonas, you know it worked out well.
How did you shoot the scene in which the Chief Elder – played by Meryl Streep – appears as a hologram?
Early on in the production, Method used The Embassy in Vancouver to create previz for this key sequence. Due to schedule issues, we knew we would not have access to both the location and Ms Streep at the same time, so we planned to combine them together later on. That meant shooting motion control, so we decided to justify the extra cost by creating more visually complex shots.
The key shot occurs when the Chief Elder appears in hologram form for the first time. We designed a move that starts off in the same volume the Chief Elder will eventually occupy, and pulls up through her body. It’s a neat effect that shows her body building up towards us as the move progresses.
Our FX team, led by Ian Farnsworth, designed the hologram look, while our set layout team, led by lead matte painter Jeremy Hoey, created the complex set plans needed to allow motion control in the two different locations. Working with General Lift, CG artist Christian Emond created camera moves that could be imported in to the motion control rig, building a hugely successful system for future motion control shots. General Lift provided the motion control system and despite our not sleeping up to, during and for a few days after, the whole thing came off without a hitch!
How long were you working on “The Giver”?
I started work on the show in May 2013, and finished in July 2014. The shoot was from September to December 2013. Turnover happened just after Christmas in 2013 through to February 2014. We had several previews, a couple of trailers and the overall VFX team hit a maximum peak of around 70 people. Matchmove and roto were done both in and out of house, and all shots were reviewed remotely over CineSync. Method Studios had regular calls with Robert Grasmere, who was in New York with Philip Noyce and the editorial team.
The Giver reminded me that details matter. The whole team at Method Studios did an amazing job of putting this show together. Our industry spends a lot of time talking about the artistry and the crazy hours spent putting those images together, but this movie was also made with the help and dedication of all those producers, digital production managers and coordinators who, after the artist team has gone home, stay to upload, download, prep, and make sure the things we spend so much time doing, actually get seen by the clients. My biggest thanks goes out to them.
Also, the on-set team in South Africa was superb. I was worried that shooting so far away would present challenges in finding on-set wranglers and coordinators, but luckily I was very wrong. And South Africa is beautiful. I mean absolutely stunning. I was amazingly lucky to spend time on its northern border, flying drones up and down Augrabies Gorge. It felt like we were on another planet.
- The Giver – official website
- Method Studios
- Mr. X. Inc.
- Mark Breakspear at IMDb
- Robert Grasmere at IMDb
- Pablo Molles at IMDb
- Philip Noyce at IMDb
Special thanks to Anthony Zwartouw, Rita Cahill, Ellen Pasternack. “The Giver” photographs copyright © The Weinstein Company 2014.