DMO: Part of the hype with Final Fantasy was how realistic the CG humans were. How far away is the industry from completely replacing a real actor with a virtual one? Virtual extras don't count…

Dave: We're getting closer all the time to matching the visual aspects, but it is the subtleties of human movement that are the main hurdle. That being said, if it hasn't been done convincingly in the next three years I'll be surprised. But I think audiences enjoy human actors for many different reasons, and I hope that never changes - but for aliens, monsters, and animals, I see that reaching even new levels of realism in the near term.

Joe: I don't think they (Square) were aiming for photo-real, and all the CG vs. Human hype seems a little over blown. I think they were just aiming away from plastic toys. A photoreal human is certainly possible. ILM did a very convincing Frankenstein test several years ago, but shelved the project. The question you have to ask yourself is, "is photoreal enough?"

Rowsby: Well, ILM tried it with their scorpion king character in "The Mummy Returns", and it's pretty clear that work between studios, or even teams in studio's, can vary greatly. I think that the tools continue to improve and so are the people using them. I would say it's around the corner.

Gregg: I don't see it happening for quite a while...

DMO: why?

Gregg: The public doesn't buy People magazine to hear that some TD upgraded 'Aki's' facial targets, but if Tom Cruise shows up at some Hollywood eatery, its big news. That's the power that causes people to pay $10 to watch a movie.

Daryl: I'd give it a few years. The real problem is if it's necessary…most of the uses for virtual humans are as extras, stunt doubles, or for a specific effect. There's no reason to have a virtual actor which takes 5-6 people to perform vs. 1

I think visually they're there; they just have to work on the motion, which is where a lot of CG fails. The mo-cap was actually about the best-acted mo-cap so far, but the rest of it was under-acted.

Mark: We are easily 5 years for a full screen, frontal close up - however, unless we make some serious headway in lip syncing it may take longer. This seems to be a real sticky point right now.

But the longer answer is 10-15 years - with a convergence of AI and CG, we may get to the next level, in my opinion, that of Directed Characters. No longer will we have to struggle over walk cycles that take a day to complete - but a director will tell the virtual character - walk over to the chair, sit down, and laugh...

GO TO PAGE [ 1, 2, 3, 4, 5, 6, 7, 8, 9, 10 Home ]