By Senior Editor Tekla S. Perry
A new motion-capture technique from the person who helped bring you QuickTime is turning actors into digital characters capable of anything imaginable.
Tekla S. Perry
I've been captured. Motion captured, that is, with one of the new types of motion-capture technology that may soon revolutionize the movie and videogame industry and make living actors indistinguishable from those that live inside computers. You can read all about this coming revolution in "Ready for Their Close-Ups" (by Eric Pavey). And you can see my digital head in action in this month's Back Story column, "Face Time".
To turn myself into a new-generation Max Headroom, I drove from my Palo Alto office to San Francisco's Dogpatch district on a rainy day in February. There, a company called Mova is putting actors in a cage and turning their performances into digital files that can be altered at a director's will, making the actors older, younger, or into space aliens.
The neighborhood, a mix of old factories, auto body shops, construction supply yards, and a power distribution substation a few blocks from the city's waterfront, isn't exactly posh. After a long climb up several flights of concrete stairs (the elevator was broken), I entered Mova's studios, where company founder Steve Perlman and several technicians and production assistants waited.
Steve Perlman? That name ought to ring a bell. Remember WebTV, one of the first companies that tried to bring entertainment content from the Internet to the home television screen. Perlman co-founded it, and Microsoft eventually bought it. How about QuickTime? Perlman led the development of the technologies that became that software within Apple, and then went on to join General Magic before founding three investor-backed companies. Now, he runs Rearden Companies, formerly Rearden Steel, a technology incubator. And Mova's new motion-capture system, called Contour, is one of the technologies Rearden incubated.
PHOTOS: MARK RICHARDS; MOVA
Given that I had to spend about an hour in make-up—Contour records and analyzes the pattern of marks made on an actor's skin when phosphorescent makeup is dabbed on with a rough, exfoliating sponge—it was nice to have Perlman there. (Click on this link to view a slideshow of the Contour process.) Because besides being an incredibly prolific entrepreneur, he is an entertaining storyteller and didn't need much prompting to tell me the story behind Mova.
Perlman always loved movies and animation. He made his first Claymation films back in junior high. He bought a motion-capture system in 1997, the kind that requires sticking little balls all over an actor and shining lights that would reflect back from the balls and be recorded as dots. He played with motion capture as more of a hobbyist than a businessman, getting more serious in 2000 when he started doing motion capture under contract to movie and game companies, eventually spinning that business out of Rearden into Mova. And he began to try to invent the next generation of motion-capture technology, one that wouldn't use markers and would be optimized for capturing facial motion and expression.
"It was an Edisonian journey," Perlman told me, reflecting on all the dead-end roads Edison took in his development of the light bulb. "We tried ultrasonics, seeing if we could measure sound waves. We tried sending light waves and trying to measure the time of flight. We tried optical flow, which is a technology that tracks pores in the face, the problem with that is actresses spend a lot of time making sure you can't see the pores. We tried retroreflective materials, but that involved tiny glass beads, which wouldn't be very safe to use on skin."
At this point, Perlman and his research team were working in an office subleased from Android, a start-up company in downtown Palo Alto. They used the central, windowless space, because, at times, they needed to be able to work in the dark. In August 2005, their officemates moved out, and another start-up, the now famous Facebook, moved in. Mova's development continued uninterrupted surrounded by the chaos of the nascent Facebook.
The group finally settled on phosphorescent Halloween makeup, typically yellow, but which, mixed with theatrical makeup, approximated a flesh tone. To give the best information on facial movement to the computer, the makeup artist had to apply the makeup to the actors face in a noticeable but random pattern. The development team tried an airbrush, which made nice spots in some areas but dense blobs in others. They tried paintbrushes, and a variety of sponges, finally settling on an exfoliating sponge. They unveiled their technology last summer, moved it up to their San Francisco studios, and are now working full out on projects for movie and videogame companies.
Back in Palo Alto, development on the next generation of Contour continues, currently in Perlman's garage. The company's lease on the space ran out at the end of last year, and Facebook booted them out. They're looking for new Palo Alto office space now.
"We like being there for a bunch of reasons," Perlman said. "Palo Alto has the northernmost Fry's. [Fry's is a chain of electronics stores that is a mecca for Silicon Valley engineers.] We rely on Fry's enormously for quick access to parts, processors, boards, and connectors. Palo Alto is also a Caltrain stop, and we have another stop here near our studio."
Perlman could have talked all day, but the makeup on my face was completely dry and I was ready to be captured. Perlman stepped aside, a director slammed a clap-board in front of my face, and it was take one. I said my piece, and, after about five takes, made a series of extreme faces; those will go to production studios as test data. Could be some famous director will decide to turn me into a strange sea creature or alien.
Newsletter Sign Up
Sign up for the Tech Alert newsletter and receive ground-breaking technology and science news from IEEE Spectrum every Thursday.
Will Overly Polite Self-Driving Cars Brake for Jerks?
Game theorist shows how pedestrians will exploit self-driving cars' built-in yen to yield
Online Voting Isn’t as Flawed as You Think—Just Ask Estonia
Casting a vote online can be secure and convenient
Swarm of Origami Robots Can Self Assemble Out of a Single Sheet
Harvard's self-folding robots are getting close to being able to walk straight out of a printer
China’s 2,000-km Quantum Link Is Almost Complete
The Beijing-Shanghai project will form the backbone of the nation’s quantum communications network
New Way to Control Heat in 2D Materials
Slipping a few lithium ions between layers of 2D materials changes how they conduct heat
Inside the Development of Light, the Tiny Digital Camera That Outperforms DSLRs
The creator of the new Light digital camera explains how he made it work
The Ever-Evolving Field of Electrical Engineering
What’s the single constant? Love of a challenge
Otto's Robotruck Hauls Budweiser While the Driver Twiddles His Thumbs in the Back
The 200-kilometer trip raises the question: when will drinking while being driven be legal?
ZCash Will Be a Truly Anonymous Blockchain-Based Currency
Set to launch Friday, ZCash is built to do what Bitcoin can't—provide privacy
Which Path to IoT Security? Government Regulation, Third-Party Verification, or Market Forces
The security flaws within the Internet of Things must be fixed, or denial-of-service attacks will only worsen
You Can't Stop Robots With Furniture Barricades Anymore
This Georgia Tech robot is smart enough to shove furniture out of the way to get where it wants to go
Pixeldelic Illuminates Silicon Valley Fashion Week
Fashion designers embrace the LED
Cypher: The Deep-Learning Software That Will Help Siri, Alexa, and Cortana Hear You
This startup pivoted from cleaning up calls to helping machines discern speech
Social Media’s Dial-Up Ancestor: The Bulletin Board System
The history of the BBS shows that pre-Internet social media was pretty great
New Record for Nanowire-Based Solar Cells Achieved
Previous record is nudged just slightly, but much higher conversion efficiencies should be achieved soon
What Is a Distributed Denial-of-Service Attack and How Did It Break Twitter?
On Friday, multiple distributed denial-of-service attacks hit a common Domain Name System provider for popular sites including Twitter and Netflix
Replace Lego’s $190 Intelligent Brick with MIT’s Scratch and a $40 Raspberry Pi
A low-cost way to program and control Lego machines
Video Friday: Robot Toes, Wolverine Haptic Hand, and More From IROS 2016
Your weekly selection of awesome robot videos
Arpa-E's $85 Million Plan to Build a Battery the Size of the Grid
Or something like that
4 Steps to Turn “Neural Dust” Into a Medical Reality
Tiny ultrasound-powered motes could record and adjust nerve activity