ADVERTISEMENT FEATURE
Meet Dorothy Clear Angle Studios tells us about their revolutionary head scanning system Dorothy: the cutting-edge tool capturing digital humans with unmatched precision O f all the challenges in VFX, creating digital humans good enough to fool real ones has long been the holy
EVERY PORE is the same , EVEN THE SUB-SURFACE BLOOD FLOW – it’s all shown exactly as it is ” heads and faces,” Ridley confirms. “We cover down to just below the shoulders, and the reason for that is to maximise the camera and lighting to produce the best images and the best scan, without blowing anyone’s eyes out with light.” The crucial difference between things and people, Ridley reflects, is practical: “you’ve got more time with a prop. You don’t necessarily get as much time with talent, so you have to maximise that time so they can get back to what they do best – acting.” Clear Angle has 18 full-body systems available worldwide, but “this rig has been designed specifically around The hardware required to make that happen quickly, as Ridley describes it, is significant, and can vary depending on the goal. “Dorothy has 1500 lights and 76 cameras that are used to create the 3D mesh. For 3D capture, we photograph
grail. At the same time, sheer demand for 3D content often tests the productivity limits of traditional modelling and texturing techniques. Meanwhile, when big-name talent is involved, the need for speed and precision has created demand for some of the most capable scanning technology ever used to capture a human likeness. Dominic Ridley is a co-founder and director of Clear Angle Studios, founded in London in 2013. It maintains facilities worldwide for scanning objects, locations and – with ever-increasing accuracy – people. Clear Angle calls its most capable scanner Dorothy; an inward- looking sphere of cameras and lights designed to capture a human head, and much more than just shape and colour. "To date, our focus has predominantly been film production, across the big studios and indies," Ridley says, "but now we’re also branching out into games and advertising. We even worked on a couple of the Super Bowl commercials this year." “Dorothy services a range of different requirements,” Ridley goes on. “From 4D to facial performance capture, to training data – we don’t use machine learning or AI, but we do capture the data. If you’re trying to faithfully recreate someone on screen, we can capture multiple lighting scenarios of someone doing a performance. If you’re doing a stunt, we can get the stunt person to come in, then the lead talent, and a VFX house could do a face replacement without having to do the old school approach of rigging and animating a face.”
around 16 frames per second. But if we’re doing 4D capture, we’ll use 6K cameras and capture 96 frames per second. It depends what you’re trying to capture.” The storm of data produced by those cameras is stored on a series of servers Ridley describes as “hefty – our server room is cooled to the max. Our electricity bill isn’t worth mentioning, but it’s a hell of a thing to be able to piece together all of this data and produce amazing results.” How Ridley often describes those results is “pore-level detail that captures high-quality textures with subtle nuances, and an understanding of how light changes across surfaces. Having pore- level detail does change how light casts across the face. You can model, say, a lampshade, but getting fine detail on a human face is difficult. That’s the stuff that’s really, really tricky to do. When big- name talent turns up, you’re so familiar
20
DEFINITIONMAGAZINE.COM
Powered by FlippingBook