Picture any Doug Liman film - Bourne Identity, Live Die Repeat, Mr. and Mrs. Smith - and you'll likely think of exciting chase scenes, set pieces filled with overwhelming, stylized action, or rapid jump cuts between scenes. Films characterized by immersive worlds, to be sure, but also by cinematic traits that might not fit the mold for virtual reality.
Yet Liman, who believes VR is "way more immersive than anything I'll ever be able to accomplish," decided to throw out the traditional rules for VR films with Invisible, a five part scripted VR series heavy on Hollywood-style action.
Read next: The best VR headsets
Liman and his 30 Ninjas studio team recently attended a VR panel run by the Tribeca Film Festival at Westfield World Trade Center in Manhattan to discuss their new series. Tribeca has created a free-to-the-public Virtual Arcade showcasing Invisible and several other high-quality virtual productions, available over the weekend through 20 November.
The panel, consisting of Liman, Emmy-nominated producer and 30 Ninjas CEO Julina Tatlock, and directors from other VR films, was in agreement that VR must evolve as a medium, asserting that most VR films are too experimental and mundane. Liman's two fundamental rules on the Invisible set were "Don't be boring" and "Don't be arty." He wanted to bring his own brand of 2D action and experience to the virtual space.
Pushing the boundaries of VR
At the beginning of their 30-day shoot, no one was entirely sure how much of Liman's Hollywood directing style would translate well into VR. Imagine how nauseous the Bourne series' notorious "Shaky cam" might make someone, for example. Tatlock discussed how they relied on static and safety shots to depict action scenes before slowly "pushing the envelope" and getting users more comfortable teleporting rapidly through their world.
"We started the project with a long list of things of what we couldn't do, and pontifications of, 'You can't do this in VR, you can't do that,'" Lewis Smithingham, president of 30 Ninjas and director of photography for Invisible, told me after the panel. Yet these rules had "exact word for word parallels" with guidelines made during "the early advent of theater and film."
In other words, 2D filmmaking remained static and limited under the regulations currently governing VR filmmaking, until visionaries like Jean-Luc Godard and the French New Wave challenged Hollywood's aversion to jump cuts and stylistic editing. So in Smithingham's mind, if these rules that dominate VR today became obsolete decades ago, "Why would they be any good now? Why limit ourselves?"
Look here, here and here
VR films typically strive to preserve the user's immersion by placing him or her inside one POV and keeping that viewpoint consistent throughout, but Liman sees this limitation as both unnecessary and selling audiences' attention spans short.
"I think your tolerance for cuts is directly related to your interest in the story," he explained. "Cuts mean you may be disoriented for a second, but then you'll be reoriented if the world is compelling." He credited Academy Award-nominated writer Melisa Wallack for keeping audiences motivated and engaged enough to still understand Invisible's world and story even when the camera angle shifted.
Relevant read: How Hollywood is starting to invade the VR film scene
In the first episode of Invisible, a fight scene breaks out in a hospital. As the fight progresses and combatants fall, the 360º camera location switches to different spots, tracking and anticipating the action.
One camera angle sees the invisible figure walking down a hallway toward a room; the next shot shifts inside that room in time to see the blurred figure step inside. The familiar sights of individual actors anchors viewers as they quickly reorient themselves back into the action from a new perspective.
Yet finding locations that fit this camera style was an ongoing challenge for the production. Liman discussed how during their first test reels, you could "watch two actors here, or you can look at a wall, or stare at the ceiling. How does that enhance the experience?" They improved the set design so locations would segment naturally into individual sections, and then used the narrative to guide restless viewers to and from each section as intended.
The team even painted walls and ceilings a certain color to emotionally manipulate the viewer when they did become distracted. Tatlock described how art director Lauren Fitzsimmons "was acutely aware of the emotional impact the color palette of the scene would have on you," whether that impact was "cold and scary" or "warm but ominous."
What makes location scouting especially difficult is finding sufficient lighting. Typical Hollywood stages will use a 50K light off-camera to light the scene, but there is no "off-camera" for Invisible. Smithingham detailed how their lighting director had to create artificial LED panels and hide as many light fixtures on set as naturally as possible for the scene to work - not always possible for less industrialized settings.
Pulled in different directions
Another tricky challenge unique to VR films: directing actors when you can't actually be on set to do so. "In any given scene, actors are used to having everyone's eyes on them, but in this case, everyone runs away, "said Tatlock, laughing," so they're alone with the camera and the other actors." This meant lots of instructing actors to memorize story beats and dialogue, then stepping back and hoping for the best.
We were looking for any technique that worked, and then experimented, and then failed.
"I had the chance to work with amazing actors that don't necessarily deliver 100% of the time," said Liman. "I'll love this and this [performance], but not this over here. So someone can choose to look over there and be disappointed. You have to look at your flaws, your warts. Everything has to be perfect. Because something I always relied on in 2D (editing to emphasize one actor's performance while hiding the other's shortcomings) was suddenly taken away from me."
In the face of so many new challenges, decades of Hollywood experience became less important than a simple willingness to try new things. "We were looking for any technique that worked, and then experimented, and then failed," said Tatlock.
Invisible's total runtime is only 30 minutes, a reasonable length for one director to handle, but Liman was "adamant" that each episode be helmed by a different director, who could bring new ideas and perspectives from which they could all learn. Simon Crane - second unit director and stunt coordinator for Star Wars: Rogue One - horror director Jerome Sable, and Michael Litwak joined Liman and Tatlock in directing an episode each.
How did this crowded hierarchy work? Tatlock admitted, "At certain points it became frustrating for certain directors because there was so much Greek Chorus of 'Will this work, will this not work?' We required the directors to be part of that conversation" and focus on technical details they might normally delegate to others. Ultimately, they meshed into a cohesive, proficient team.
The medium is so visually compelling that you really have to lay out the story in a simple way...
Much of the experimentation started with the cinematography. DP Smithingham sought inspiration not from action films, but from classic films that subverted the conventions of their day: he cited D. W. Griffith's Intolerance for editing techniques, Godard's Breathless for jump cuts, and especially Hitchcock's camera movements as models for working in VR. "I think it's really important, when you're creating a new medium, to rely on history and look closely at what succeeded or failed," he said.
Tatlock jumped in there to note, "I challenge you to find a Doug Liman production where, for an action film, one of the team members brings an experimental film like La Jetée to the table and say, "It'll be like this!" and have the team accept this." She also emphasized that Liman himself "is not a rear view kind of person" and focused more on bringing his own style to VR than looking to history for inspiration.
For all of this experimentation with style, the team did their best to keep the substance as simple and straightforward as possible. Tom Vance, head of content at Jaunt VR and a producer on the series said, "The medium is so visually compelling that you really have to lay out the story in a simple way, so they can follow the story without getting distracted by what they're watching. We went through a number of iterations around a number of complex and ambitious storylines and ended up reducing it to the propulsive bare bones that would keep the audience engaged."
And that's a wrap
By the end of their packed month-long shoot, the team had drastically improved its shooting ratio and competency with VR cinematography. Everyone from Liman downwards appeared very excited about jumping right back into a second season of Invisible, so that they could take what they've learned and generate an even better product.
Yet when it came to the question of monetizing their work or quantifying Invisible's success, things remained speculative. Liman demurred that "I'm in my own little cave as a storyteller; the reason why I live in New York instead of Hollywood is because I want to focus on my own work, not the future of the business." By contrast, Tatlock noted that VR budgets "are closely aligned with a high quality cable show. That's a lot for an emerging medium," acknowledging that VR can't stay in the demo phase forever.
Vance doesn't take a doom-and-gloom approach to this problem. "I'll reference an article from Newsweek from '95 about the internet as an isolating thing where no one knew how to make money off of it. All those same questions are being thrown our way (about VR), and I find that really inspiring. That (money) problem is something we think others will solve, so we just want to set the table to make great content and engage an audience and make the case for monetization down the road."
Invisible is available for free now on the Jaunt VR and Samsung VR apps through Gear VR, Oculus Rift, HTC Vive, iOS and Android, as well as desktop.
How we test