to join the CVG community. Not a member yet? Join now!
CVG
News

David Cage: L.A Noire tech is 'interesting dead end', Quantic's new approach is the future

Photo-realistic real-time graphics here in five years with performance capture, says Heavy Rain man

Heavy Rain creator David Cage has predicted that photo-realistic graphics will be possible within the next six years - and revealed that he is unconvinced by the long-term potential of new facial capture techniques used to make Rockstar's L.A Noire.

Speaking to CVG, Cage said that Quantic Dream is already employing new technology to capture full body performance in unannounced games. He claimed these projects were already showing "significant progress" on what the studio achieved with Heavy Rain last year.

L.A Noire benefits from DepthAnalysis' Motion Scan technology, which requires in-game actors re-record scenes whilst just their head is filmed by 32 HD cameras from various angles. This has allowed developer Team Bondi to represent subtle facial movements during gameplay. Bondi and L.A Noire co-creator Rockstar have proudly claimed that MotionScan represents a "revolution" for motion capture in games, but Cage is less convinced.

Zoom

He told CVG: "What to say about L.A Noire? I think it's an interesting solution to a problem for now. But it's also an interesting dead end. That's exactly what I feel. Their technique is incredibly expensive and they will never be able to shoot body and face at the same time."

Cage revealed that Quantic Dream is using performance capture that reads both actors' body and faces in one shot - which he believes will lead to real graphical advancement.

"We are doing that now [at Quantic], and our next games will be shot with performance capture," he explained. "We see a huge difference between shooting the face and body separately and shooting everything at the same time. Suddenly you've got a real sense of acting that is consistent. You can't imagine how related what you say with your face is to what your body does.

"[Those using MotionScan] will never be able to do that. The other thing is that they can't have real time lighting. Their technique means they can't have lighting the way I think we should do it. Basically, they take pictures; they take scans several times per frame. They also have limitations on the shaders they use, they can't re-target the eyes because they eyes are captured. When you have actors in real time you like to to re-target the eyes to make sure they look at each other [convincingly] etc. etc.

This video is no longer available

"It's a list of important problems that cannot be solved with their technology. I think L.A Noire looks good - honestly, it does - but I don't think they'll go much further than where they are. With the technology we use, we can improve; there is a lot of room for improvement and we hope to show very soon where we are now. We've made significant progress since Heavy Rain and will continue to make progress until we reach the stage of Avatar. That is probably three, four five years from now."

Zoom

When we asked if Cage was referring to the photo-realistic visuals seen in James Cameron's blockbuster movie, he replied:

"Yes. Four or five years. Maybe it's going to require a new [gaming] platform, but when you look at where real-time is right now, it's probably where CG was five, six years ago. You can say there's a five-year gap between CG and real time. Avatar was released last year, so that's where we should be in five, six years from now."

Rockstar this week revealed that L.A Noire will be the first video game to ever feature at the respected Tribeca Film Festival in New York next month. The detective thriller will be released on 360 and PS3 in May.

Comments