THE MILL
If Jaguar’s new The Lab commercial seems cinematic in its look and feel, that’s because it was shot on the 007 Stage at Pinewood Studios in the UK. The spot opens on a feature-film scale in a lab, where the “brains” behind the operation (played by actor Nicholas Hoult) takes viewers, quite literally, into the depths of the innovation lab where Jaguar fuels the great British villains seen in the automaker’s 2014 Super Bowl spot starring Sir Ben Kingsley and Mark Strong. After all, “behind every cunning plan is a brilliant mind,” says Hoult.
The Mill in London (www.themill.com) created VFX for The Lab, delivering four 30-second versions, a 90-second digital film plus online and print stills. Academy Award-winner Tom Hooper of Smuggler London did the directing for agency Spark 44 using two Arri Alexa cameras and three Canon 5Ds to capture the expansive live action.
In one sequence of the spot, Hoult rides in a transparent elevator tube, which magically appears in the empty room of a manor house. It travels down through a huge swimming pool, replete with circling shark, past multiple garage levels to arrive at the cavernous lab space.
Hoult wasn’t available for the live-action shoot at Queen’s House, a former royal residence in Greenwich, so he was shot greenscreen at Pinewood, while a small crew captured the backgrounds at Greenwich along with a plate spanning the width of the pool.
“Once we went underwater, things got a little more tricky,” says The Mill’s 3D lead artist, Mike Chapman. “The initial camera move that was shot on a crane with Nicholas’ help was static, so we needed to track this plate and then take the vertical move off the camera and put in onto the lift [elevator]. This was all carried out before we could even think about the background and the shark, which were both full CG.”
The “heritage garage area” that Hoult passes through showcases Jaguars from different eras — even some new ones still under wraps. “The space was created fully in CG, starting with our concept team working closely with the 3D team to design and build different variations of the space based on some of the design details we found both in the 007 Stage at Pinewood and in Aldwych tube [subway] station,” Chapman explains. “Once we had the space built, we tracked Nicholas’ action from the greenscreen at Pinewood and comp’d him into the lift.”
The central portion of the lab is full of floating CG car parts made from CAD files supplied by Jaguar. Hoult was shot on a cast acrylic platform built on a rig in the center of the 007 stage, which enabled Hooper to keep the background in camera along with reflections on the floor.
“Then we added what you would see refracted underneath along with the rest of the glass box, tube tunnels, and floating car parts,” says Chapman. Hoult interacts with a few car components, grabbing one in the air and pressing the button on another. “As far as the interaction goes, it’s a combination of real car parts puppeteered for the close-ups alongside animated 3D parts in the wide shots,” he explains. “We did have a bunch of the parts on set for reference, so we took a lot of photo reference and did some reference passes on the stage so we could faithfully match the lighting in CG.”
One particularly clever shot has a Jaguar’s floating aluminum chassis passing right over and through the actor. “We rendered a bunch of parts of the chassis separately so we could flicker them on and off in 2D along with 3D rooting Nicholas to get glow for the interaction points back on him,” Chapman says.
Autodesk Maya was the main 3D package for modeling, animation, and lighting with The Foundry’s Mari and Adobe’s Photoshop for texturing. A combination of Pixel Farm’s PFTrack and Science-D-Visions’ 3DEqualizer were the tracking tools, with Solid Angle’s Arnold for rendering. All compositing was done in The Foundry’s Nuke and Autodesk’s Flame.
“We have a proprietary asset management program built into Maya, which was really important for us when managing so many individual elements, like all the cars,” notes Chapman. “We also have an in-house procedural rendering workflow, which basically means we don’t carry any of the heavy CAD geometry in our render scenes and just load these when we need to render. There were no tools specifically written for this project, but we did come up with quite a nice rig for scattering the car parts.”
METHOD STUDIOS
Spot watchers are seeing a lot of Rob Lowe these days. In fact, the actor stars opposite himself, playing a bizarre array of Rob Lowes in a DirecTV campaign from Grey New York. Directed by Tom Kuntz of MJZ, the comedy spots depict handsome Rob Lowe enjoying the benefits of DirecTV, while Creepy, Hairy, Painfully Awkward, Less Attractive, and Scrawny Arms Rob Lowe suffers through the indignities of being a cable customer.
Method Studios (www.methodstudios.com), which is headquartered in Los Angeles and has eight other offices worldwide, did the campaign’s VFX. In addition to the split-screen work, which enables Lowe to interact with his other persona, Method handled prosthetic cleanups, digital matte paintings, 2D enhancements, and compositing. Most of the spots rely on Lowe’s performances and prosthetic makeup to create alternate versions of the actor. For instance, Scrawny Arms required Method to create much of the other character, who has skinny, muscle-less arms that can’t even twist the lid off a mayonnaise jar.
“At first we thought we’d do head or arm replacements,” says Method VFX Supervisor Jay Hawkins. “We did some proofs of concept shooting a skinny girl and a muscular dude in a few poses and doing some rough composites, and they were really creepy. It was the muscle movements: If it all wasn’t connected, it looked scary — and not in a funny way.”
Method was inspired by the first Captain America feature, which involved “skinnying up” the character before he becomes a superhero. “It was a very elaborate process,” says Hawkins. “We didn’t have the budget or time to render a 3D human for the spot, but was there a way to do it in 2D?”
They did another test, placing tracking dots on the arms of the muscular fellow, then removing his arms, skinnying them up, and putting them back, segment by segment, to create a 2D puppet. “We could squeeze him, expand him, twist him, and keep the pivot points locked together,” Hawkins explains. “The test was super successful, and everyone was on board with the process.”
So Method shot Lowe on the set with tracking markers on his arms; they also shot a body double for muscle reference and framing. Then they removed Lowe’s arms and put them back, applying precise and unique warping for each segment.
The only 3D element was his watch, which dangles loosely from his wrist when Lowe, sitting on a couch, raises his arms and locks his hands behind his head. “It was the most difficult shot,” Hawkins says, “because so much was connected — the movement of his biceps, armpits, and chest.”
Flame was Method’s chief VFX tool; the watch was modeled and animated in Maya. Nuke was deployed in other spots for prosthetic cleanups. Carlos Herrera was the VFX producer for Method Studios in New York.
PSYOP
Hey, Gen X-ers! Remember the feeling you got when you found Skeletor and He-Man, Jem, GI Joe, Strawberry Shortcake, Stretch Armstrong, and Gumby and Pokey among your holiday presents? Honda wanted you to recapture that feeling once again with a slightly higher-priced “toy” in its A Gift to Remember campaign from agency RPA.
Bicoastal Psyop (www.psyop.tv) teamed with LA’s Screen Novelties stop-motion studio to create the nostalgia-filled TV spots (one commercial, featuring Fisher Price's Little People, spans several generations), as well as an “extras” for social media. Psyop’s Fletcher Moules and Todd Mueller directed, using the toys to illustrate various attributes of the cars (like rear-view safety cameras) in a homey car showroom.
The toughest part of the campaign was tearing the directors away from playtime with the vintage toys purchased for the spots. “I got to animate He-Man and Skeletor, just like I did as a kid with my VHS recorder: record, pause, record, pause,” laughs Moules. In pre-production, he and Mueller spent a day with a camera, a Honda Odyssey, and all the toys, “playing like 10-year-olds” to “discover how to tell stories inside a [real] car.”
From the start, they felt strongly about using actual vintage toys, not photoreal 3D models, to convey the sense of nostalgia at the heart of the spots. The only exception was the Little People spot, which required hoards of the tiny cylindrical toys, best executed in 3D.
All of the toys were cut open to accommodate wire armatures for stop-motion animation; Stretch Armstrong was also recast, molded, and remade for one sequence where he needed to reach from the backseat to the dashboard.
The car exteriors were shot with Arri Alexa in an empty dealership dressed like a living room. Then the directors moved to Screen Novelties to shoot stop motion with a Canon EOS 5D for almost three weeks. “We had two stages with real cars and motion control,” says Moules. “The cars were up on jacks. We had to work around the physicality of the vehicles, removing certain doors and seats and the windshields so we could get in the cars with the toys and not be hindered by the camera or motion-control arm.”
Only a few shots were greenscreen because of the camera angles they wanted to capture — like the shot of Gumby and Pokey peering down the cup holder as if it were a cave.
The fun scenarios, developed by RPA and Psyop, matched the personalities of the toys. One of Moules’ favorites has Gumby and Pokey repelling down the dashboard with the same sense of adventure they used to get from crossing from one end of the living room to the other, he notes. Stretch Armstrong is revealed coming out of the trunk; GI Joe tosses a grappling hook over the rear-view mirror — a disco mirror ball on the other end. Jem’s hair blows in the breeze of the vent.
Gumby had a wardrobe of face parts, “hand-cut, sticker shapes to remove and replace to give him expressions,” Moules explains. Most other moving toy mouths were done in CG. “For Stretch and Jem, we modeled their faces in CG, lip-synched them in CG, then composited the CG face from the nose down onto the stop-motion characters.”
The Little People, who assume a tower formation so they can look through the car windows, were modeled and animated in Maya and rendered in Arnold. “We shot a plate of the real toys — with us holding them in place — on set for lighting and scale references,” says Moules. The dealership settings were composited with Nuke and Flame.
BREWSTER PARSONS
Fire King Vargas and Lava square off for a sword fight, and he looks cocky about besting her in one-on-one combat, until the camera pulls back to reveal a vast army silhouetted behind Lava. “It’s good to have friends,” says the voiceover in the animated spot for Gumi’s Brave Frontier, a classic role-playing game for Androids, iPhones, and tablets.
Brewster Parsons (www.brewsterparsons.com) in Venice, CA, was tasked by Gumi and agency Ignited with giving a “Western Anime” look to the animation for Brave Frontier: One on One + 500, using the game design as a jumping-off point. “Western anime” would give the characters a more elongated, sinewy, and grittier look without losing the specifics that make each character unique.
“Gumi loves all the classic ’90s anime,”
says TJ Burke, creative director and CG supervisor at Brewster Parsons. “The trick was to match that style with our 3D tools. The look of anime is not physical reality. And 3D tools are made to build what’s physically real. So you have to break out of where the math wants to take you and add more of a hand-built feel.”
Burke discovered that initial notions of physically-based effects, such as simulated hair, cloth, and fire, “looked too perfect. No one responded to it well. So we went on to stylized, texture-based or hand-animated approaches.”
Brewster Parsons began with storyboards, which they cut into an animatic then built into a previs. “It was nice that we kept evolving the same setups on the previs we started with,” Burke explains. “We did a lot of tweaking with the shaders, and we built tools for automating the pipeline and building render passes, which we combined in ways we hadn’t expected.”
The company used Maya for modeling and animation, Chaos Group’s V-Ray for rendering and Nuke for compositing.
“We were in the design phase almost to the end of the project,” Burke notes. “Getting the right balance of flatness, dimensionality, and outlines all depended upon the game characters. Hitting ‘render’ was pretty easy, and revisions were super fast, but getting the proportions and design right were the big challenges.”
Christine Bunish is a veteran writer and editor for the film and video industry. She can be reached at
cbunish@gmail.com
.