VR Cinema: Keep Trying

After two and a half nonstop hours of watching VR "cinema" this weekend, I reached two conclusions:
--My head hurt, and
--This stuff is not yet ready for prime time

The setting was Cinequest, Silicon Valley's quirky independent film festival. This year it added a VR "experience," with eight half-hour VR programs you could watch, at ten bucks a pop. I chose five of them. They were a diverse selection: Big-budget Hollywood movie excerpts done up in VR, independent animation, what appeared to be game trailers, and some live action shorts.

I came in with high expectations: I've always been fascinated by 3D computing, and my first experience with an Oculus Rift was close to a religious event. So I was excited to see Cinequest's "new and amazing worlds" in which "you don't just watch, you actually experience these movies all around you," as the program put it.

Cinequest is a cool organization and they put on a great show. They're a nonprofit, staffed heavily by volunteers, and I applaud them for trying this experiment. But mostly what the VR experience showed is that our technology, and VR cinema itself, isn't yet living up to the hype.

That's not too surprising – we're still in the very early days of this new platform, and my experience with every new platform is that you get a lot of weird experiments while people work out what they can do with it. Based on what I saw at Cinequest, VR cinema is still in the weird stage. Below I'll give you details on each of the shorts I experienced, but here's a summary:

The technology needs more work. When you came into the VR room, the staffers equipped you with a Samsung Gear VR headset with a Galaxy smartphone and a pair of wireless earphones, and told you how everything worked. So right off the bat, this wasn't a movie-like experience; you don't just sit down and watch. The staffers did a very good job of teaching people and maintaining the devices (more on that below), but it was still confusing. The most puzzling part was that there were volume controls on both the headset and the earphones, and you had to turn them both to max in order to hear the content.

The on-screen interface was familiar because I'd played with a Rift before, but as soon as I started my first program I had problems. The video was running at about five frames a second, and the sound seemed way out of sync with the images. After several minutes of futzing around with the controls, I gave up and called over one of the staffers. He explained that the Galaxy smartphones used in the headsets were getting overloaded by all the video files, and had to be restarted regularly. He rebooted my system, a procedure I had to do two more times in the two and a half hours.

Now the video was running at good speed, and I was very pleased that I didn't experience any lag when I moved my head. But the images were grainy, far more so than either a film or television show. The color palette seemed to be limited as well – the live action videos looked washed out, peoples' faces were monochrome, and in dark scenes there was noticeable pixelation. It reminded me of watching an old pre-hi-def color TV.

None of the programs were as immersive as a good movie. In movies we have almost a century of experience in how to tell a story visually. VR is different enough that we need a new set of best practices. For example:

--The camera was sometimes in odd positions. In one film, you appear to be sitting in the passenger seat of a car, but squashed down about a foot above the seat so you're looking up at the characters and can't see out of the front of the car. Instead you have a panoramic view of the world's largest car stereo.

--You don't know where to look. In some of the films I ended up looking in the wrong direction and missed important action.

--Whiplash. One of the films featured a tense discussion between two actors, one on your right and one on your left. You had to whip your head back and forth to follow their interaction. That got old really fast.

--The seams get in the way. Live action VR is filmed with multiple cameras pointing in different directions. The edges between the camera images are blended so you don't usually notice them. But occasionally a character would step into the border between them and his head or some other important body part would disappear.

--It's hard to do closeups. There's a very fine art to the way a film communicates human interaction, a subtle rhythm of closeups, reaction shots, etc. A VR film can't jump your perspective around that way – you'd feel like you're being teleported all over the room. So your perspective tends to stay in one or two places for the duration of a scene, which makes it feel a bit like watching surveillance camera footage. Instead of being in the story, you feel like you're spying on it.

Add these issues to the resolution and color problems, and often I found myself paying more attention to the technology than to the story.

The rules of storytelling still apply. In some of the films, the script and storytelling were awful. No amount of great technology can compensate for awkward dialog and a lack of conflict. Ironically, the worst offender in this area was one of the big-budget Hollywood productions. You'd think they would know better.

I doubt that cinema is the killer app for VR. Even if all of the problems above were solved, I came away doubting that cinema experiences will be the thing that pushes VR into the mainstream. For me, the thing that makes VR special is its eerie sense of presence, the feeling that you're actually in another place even though you know you're not. The VR films gave me almost no sense of presence, which surprised me. I felt like I was in a wraparound Imax theater (with bad image quality), rather than being transported to a different place.

I think the problem is that in a movie your point of view has to be controlled in order to tell you a story. The movie pushes you around – sometimes gently, sometimes forcefully, but almost always you have no control. I think the ability to move around is an important part of the sense of presence in VR. Without it, the whole experience was much less compelling. I think I'd prefer to watch a conventional movie; the resolution is better, and you don't get a headache from the headset shoving your glasses into your face.

What it means: Keep looking

VR today reminds me of the early days of multimedia: We're seeing some interesting bits and pieces, but they're more like curiosities than finished products. I think we'll need a lot more experimentation, and better hardware, before VR will be ready to take off in the mainstream.

Multimedia software came of age in 1993 when Cyan released Myst, the first software title to fluidly merge the large storage of CD-ROMs with high-quality graphics, sound, and interesting experiences. Along with a couple of other popular titles, it created a whole multimedia industry in the 1990s. If we've found the Myst-equivalent for VR, I didn't see it at Cinequest.

Details on the programs

Speed Kills. VR scenes from an upcoming movie starring John Travolta. A movie about speedboats and drug runners ought to be gripping in VR, but this was the weakest program of the bunch. The scenes (which didn't fit together into a narrative whole) were mostly tedious: Travolta feeding a horse, Travolta hitting on a waitress, Travolta talking to a guy in a restaurant. To make it worse, they inserted credits and titles between every scene. So the whole thing felt like a bad commercial.

La Camila. This is a cute animated story with lovely colors, and it was obviously a work of love for the people involved. Unfortunately, the character models were surprisingly primitive. My expectations have been skewed by Pixar, and it was jarring to see people and animals that look like a bunch of linked sausages bouncing like marionettes. Unfortunately, about 2/3 of the way through the program I accidently restarted it while adjusting my headset, and I couldn't get the video to fast forward to where I'd been. So I moved on.

The Humanity Bureau. More movie excerpts, these from an upcoming Nicolas Cage movie. Much better structured than Speed Kills, but I was distracted by some very strange camera angles. There were some good outdoor sequences, but when the story moved indoors I felt the surveillance camera effect very strongly.

The Recall. A VR experience based on a 2017 alien abduction film featuring Wesley Snipes. Stilted and confusing. It reminded me of the Geico commercial parodying horror films. This is the one where I missed a lot of the action because I was looking in the wrong direction, but the things I did see were unintentionally amusing rather than scary.

Boxes. Much better thought-out than the movie excerpts, this is a live action short in which a young man cleans out the home of his late parents, and reminisces about his childhood in a series of flashbacks. A nice story well told, but I don't think it gained much from the VR.

Volt: Chain City. A frantic four minute animated chase with Star Wars-style speeders plunging through a landscape of wreckage. Hello motion sickness.

Women on the Move. A sweet live-action story about a woman in Niger who has high hopes for her granddaughter. It was an interesting visit to a village in Africa, and the VR did give me a good view of the homes and streets of the village. But I didn't feel like I was there, probably because I couldn't move around on my own.

Doctor X: Pale Dawn. Dinosaurs chasing a dune buggy. Even more dizzying than Volt.

Hutong in Live. A love letter to the Hutong lifestyle in Beijing, this one was interesting because it mixed animation and video. Unfortunately, the animation was very limited – the models were low res, you could only move between predetermined spots, and your perspective jumped from place to place rather than moving smoothly. Other than the 3D, it reminded me of a QuickTime title from 1992. I think it would have been much more successful if it had recreated a hutong and allowed you to move through it freely.

Meeting Rembrandt: Master of Reality. An animated interaction with Rembrandt. Nice idea but not very engaging. It felt like an explanatory video you'd see in a museum.

Ultraman Zero VR. Campy but fun: A guy in a monster suit attacks a scale model of Tokyo, and is defeated by a guy in a superhero suit. It was kind of fun to be between the monster and giant superhero, with both of them towering over me. But they still looked like a couple of guys in suits, and the novelty wore off quickly. If I were an Ultraman aficionado I probably would have been more charmed.

What do you think? Have I missed the point? Is there a killer title I should have watched? I'm interested in your comments.


Anonymous said...

I had much the same feeling with some Samsung VR experiences at MWC!

I think the technology will get there, at least for standing somewhere artificial and looking around... still not sure how walking around and exploring is going to work, though. So if you want to see somewhere you're never likely to travel to, like the deep ocean, the surface of the Moon or a Mars colony, or Petra in it's heyday, then it'll be fantastic soon.

You hit the nail on the head, though: it's the narrative and the conventions of storytelling that need work. So much work. All of our media create immersive worlds, engaging characters and compelling stories. Irrespective of how close to "reality" they are. Letters on a page turn out to be pretty good at it, as are comic books, concept albums, computer games, TV shows and cinema. The mind is wonderful at turning content at any level of abstraction into an experience. In some ways, the more "realistic" the content is, the harder is is to create really compelling storytelling, it seems: Blade Runner 2049 was the first film I've seen that I think was better in iMax than in a regular cinema.

And all of these media have storytelling devices and conventions that their makers use to sell the story and exploit and play with to powerful effect. Scott McLoud's "Understanding Comics" is a great read in that respect.

What VR needs are some pioneers who can work out how storytelling works in VR. Whether they'll come from the movies or gaming industry, or somewhere else entirely, who knows? Hopefully, when they arrive it won't take too long: pretty much every trick in movie storytelling is used in "Wings," from 1927.


Michael Mace said...

Thanks, Matt. I agree very strongly about tourism and VR. And I'm glad to hear you had a similar experience at MWC. So it's not just me.

Comics are a really interesting analogy. I need to think about that some more.

For moving around, I think some sort of handheld puck will be adequate. I didn't love the one that came with Rift, but it was serviceable. What I don't need is an infinite treadmill so I can move using my feet. To me that's more about simulation than immersion.

Regarding storytelling in VR, I'm starting to suspect that we'll need to revisit what we mean by "story." If a sense of presence is what's truly special about VR, then it's better suited to saying "here's what it's like to be here" rather than "let me tell you a linear story."

Maybe photography is the right way to think about it. You can tell a story through photos, but it's more a set of feelings than a straight narrative. Poetry rather than prose...

Unknown said...

LiquidCinemaVR.com addresses many of the narrative issues you’ve mentioned.