Slashdot Mirror
Digital Replicas May Change Games and Film
Carl Bialik from WSJ writes "Steve Perlman is touting technology that he says can create animated digital reproductions of the human body that are as accurate as photographs, the Wall Street Journal reports. From the article: 'Game makers could use the system, called Contour, to create very realistic animated characters in videogames with fully controllable movements and facial expressions. Film makers could use the technology as a kind of digital makeup, changing an actor's looks or words or switch camera angles without costly retakes. The technology can even substitute one actor's face for another's and create exact replicas of long-dead historical figures.'"
Given that TV studios already like reality TV in large part because the cast is cheap, will we start seeing 100% virtual actors? From a business standpoint, intellectual property beats a human face that ages, gets into tabloids ( and potentially ruining the carefully marketed image ) is costly and needs to be recycled regularly.
This may revolutionize the porn industry. Imagine taking all the best porn actresses in their primes and putting them all in one movie. Check that, 1,000 different movies. Now it's possible. On a serious note, the less actual sex involved in making the porn, the less risk to the actors.
Leaving the p0rn industry for a moment, anyone who ever sold a picture of themselves and waived all future rights and royalties is going to be in for a surprise, especially if the picture is them in a birthday suit. You may see congressional action to protect people from having their images used in such ways if they signed "all rights" contracts before the technology became available.
Someday, we will have the ability to create totally new "people" for movies, without relying on any existing images. That way the whole concept of royalties and rights is avoided altogether.
Knowledge is how to play a game, intelligence is how to win, wisdom is knowing what game to play.
And to deal with your second question, textures will still be added after capturing structure and movement information, so you could make people blue, tweak the models or whatever after the fact still. (Like in the example, the teeth, skin, hair etc are all being slotted back in from photo's)
Did you go near the article? Yes yes, I know, I must be new here...
I bet Homeland Security will love this.
Suspect that someone is a terrorist, but have no evidence at all to support your allegations. - No problem, just whip up a photo-realistic animation of them attending a local bomb-making class. Lather, rinse, repeat.
create exact replicas of long-dead historical figures
I don't think you want to witness this, judging by how the process is described in TFA:
First, an actor's face is coated in ordinary phosphorescent makeup like that worn by children at Halloween. The actors then conduct their performance in a studio surrounded by fluorescent lights and digital cameras.
Dig 'em up, cover 'em in phosphorescent makeup and dangle 'em in front of the camera?
Game makers might achieve photorealistic representations of human appearance and motion, but our new (mostly welcome) digital overlords will still bump into walls, get stuck behind things, get in your way, not look at you while they're talking and generally make mistake and act like they're just computer representations. Game makers for the most part have all the graphical juice they need to convince us of a world's authenticity.
Though I really do enjoy advances in the level of graphical detail that increasing sophistication in hardware and software bring, I feel we need better AI, not fancier graphics . If a game's AI was as big a selling point (and therefore had the same amount of money invested in developing technologies and software for it's advancement) as the graphical prowress of the hardware then I think Alyx in Half Life 2 would probably have gone sentient at this stage.
"I'm sorry Gordon, but your apparent lack of regard for your own safety means I don't want to get involved with you, i'm just afraid of getting hurt."
When the posters fear their moderators, there is tyranny; when the moderators fears the posters, there is liberty.
Given that the top stars regualarly charge $20million for a movie, that actually is a bargain.
The challenge, though, would be to get your virtual actor to that star status (so much harder without the chat shows and celebrity magazine stories about who he might marry).
I think in some aspects this is bad news for films. I'm completely turned off the idea of watching a movie if I think that they've messed about with an actor's face in order to improve on their expression or fix something. I think it's horrendous. I wonder how many actors will shortly have in their contract that film makers can't animate over their face without written consent? Speaking of which, I wonder how many actors currently have written into their contracts/estates that their image can't be raped after their death in cheesy car adverts etc?
But that's $2000/second this year.
Next year it will be $1000/second. The year after, even less. In 5-10 years, it will be possible to do it on a $2000 desktop in near realtime.
You just have to look at how quickly morphing technology went from top blockbusters only (Terminator 2) to TV commercials, to something that can be done at home right now.
Yeah, it sounds sort of like chroma keying turned on its head. Rather than shooting the actor against a background that's easy to remove, you just paint the parts that you do want to show up with the fluorescent paint. I'd guess that this would make shooting in an ad hoc environment much easier. No need to set up a soundstage with a green screen, just set up the special lighting rig wherever you want.
Definitely a clever and newsworthy idea.
Now I see some are afraid of forged "photos".... Actually, the larger problem with this technology is not that you could possibly manufacture photographic evidence to damage someone else's ambitions, goals, or whatnot; it's that in creating that possibility, hard photo/video evidence loses its credibility in court, and will only continue to do so with time.
I guarantee that when this becomes mainstream (just as most CG geeks knew would happen years ago), that implicating a person of influence/wealth will become nearly impossible, as any time any damaging photo/video evidence pops up (oh, say, like photos of torture at the hands of the US government at Guantanamo or a worse and nameless fascimile) the powerful will declare that it's been manufactured by the opposing side.
Ex nihilo nihil fit.
"Someday, we will have the ability to create totally new "people" for movies, without relying on any existing images."
s m.movie.html
This has been done in Japan. The first one I remember is Kyoko http://www.cnn.com/TECH/9702/04/japan.date/kyoko.
but I've seen other, more realistic stuff since then.
How long before (virtual) snuff films are so real the "thought police" legislate against them?
"The mind works quicker than you think!"
This may revolutionize the porn industry. Imagine taking all the best porn actresses in their primes and putting them all in one movie. Check that, 1,000 different movies. Now it's possible. On a serious note, the less actual sex involved in making the porn, the less risk to the actors.
If this technology is used for pr0n its only a matter of time until someone sells custom-made pornos of the buyer with the partner(s) of thier choice.
Blessed are the 1337, for they shall pwn the earth.
Does anyone else here cringe when they see computer animation superimposed on film or passed off as a real action sequence? Graphics have definitely gotten a lot better over the years but I always laugh when someone describes the next gen of movie effects as lifelike or realistic.
Take Spiderman for example - I find those swinging sequences to look so horribly fake and robotic. The character model looks pasted-on because the light doesn't strike his body as it should and he isn't as softened by the camera as other objects, and the body motion appears jerky and forced.
So we'll see what the next gen has in store for us, but I have a feeling it will impress us only the first time we see it in a theatre. Jurassic park looked amazing that first time, but in subsequent viewings anyone can easily tell the difference between CGI and a model.
I love video games but I hate movies full of computer effects. Practical effects like those in Sam Raimi's movies are still the only way to go in my books!
The /. comment is misleading. The technique to digitally capture the surface of a human body will not help make digital movies with no actors. You still need a real actor to do the job; the described technique only projects the original actor's image on the new actor. The age of the digital actor is not here yet, although this technique may be useful for ressurrecting dead actors.
In order to make fully digital actors, there are several problems to be solved:
1) animation that follows real life physics. Although digital animation has made great steps, the human motion can not be fully synthesized yet in a way that it totally fools the eye.
2) realistic voice synthesis. Computers still can not make realistic synthetic voice.
3) putting emotion into the above. In some day truly believable synthetic animation and voice will be achieved using only digital techniques. But what about the emotions? humans can do many emotions at the same time, all with subtle expressions, and using their face and voice in various subtle manners.
I would love to have truly believable synthespians. It would allow my favorite series to keep going on for ever. For example, there would be no problem doing a new Star Trek movie with Kirk and Spock (many fans have disagreed with the new movie due to the new actors that will be playing Kirk and Spock). But I just don't see it in the near future.
$3.6m to get rid of the real Tom Cruise sounds pretty reasonable to me. Shame it's only for 30 minutes though.
She was apparently a moderate success, a typical idol. Not all idols are shortlived but she was and with so many real girls wanting to be idols who wants to create a virtual one? They are so hard to audition on the couch if you know what I mean.
But yeah, you can see the appeal of a virtual Han Solo or Indiana Jones. Just crank them movies out without having to deal with a grandpa actor.
MMO Quests are like orgasms:
You may solo them, I prefer them in a group.
Why is it that each time someone claims to be capable of creating "photo realistic" or "lifelike" digital images only have a shity demo to show what might pass as "realistic" to the clinicly blind?
Does the film, "The Running Man" ring a bell? And if so, do you remember how technology that was exactly like this, was used? Scary to say in the least...but I am cynic by nature so...
I'm not sure I agree with your thoughts on it helping "small production films."
Actors are cheap; CG is expensive. The percieved 'cost' of actors is distorted by how much it costs to hire a super-star, but most low-budget films can't afford that anyway, so they're using no-name actors to begin with. I think the actors' salaries are a pretty small part of most small-budget films who aren't trying to hire someone with name recognition.
A machine would definitely be cheaper than hiring Harrison Ford, but to paraphrase Monty Burns, it's probably not going to be cheaper than hiring his 'Mexican, non-union replacement.'
In any major city there is probably a surplus of actors willing to work in film for a basic living wage (and quite possibly less than that). Particularly if you can script your film to use youngish actors, there's a fairly big talent pool of people willing to work for publicity; in some cases they can be quite good.
"Ladies and gentlemen, my killbot features Lotus Notes and a machine gun. It is the finest available."
"Realistic animations are already possible, has been for ages, it's called motion capture."
Motion capture doesn't work for the face. You could be thinking about performance capture, which does capture the face, but there's debate about how effective that really is. In any event, no, it has NOT been here for ages. If one actor really can drive the actions of another, this is a Big Deal TM. You would not believe the amount of work that is done to deal with facial movements on a character. Check out the extras DVDs on King Kong or I, Robot if you're really curious about it.
"I only see the use of this technology for movie to game adaptions were they can quickly copy a real life actor to 3D. For the rest, why would you want to hire multiple actors to do the same thing what a couple of voice actors, motion capture actors and animators can do."
Funny, the article had a couple of interesting ideas in that department. The character aging in reverse gag, for example, is a rather interesting one. As for the latter half of your question, the answer is time. The end result is a moving character. It's time consuming to hand-key animation, not to mention the potential for lack of subtlety. If you can just throw one talented actor into a scanner and get the performance you need with minimal clean-up, you're in a better place.
"Besides, how would you use this technology in a non-realistic game."
Have you played San Andreas?
I realize a lot of people in this thread don't see the point. Just remember that the human body is the hardest thing to get right when it comes to CG. Remember all those complaints in the Star Wars prequels and the Matrix Trilogy about the digital doubles not looking right? Contour may or may not drag us from that rut, I couldn't tell you. What I can tell you is that it's still a problem today and it's a Good Thing if they can find a solution that allows the talents of actors to drive the performance of a CG character. The possibilities are a lot broader than a lot of you can imagine. Go read an issue or two of Cinefex. You'll be surprised at what technologies are already making a huge difference in modern movies, even though you probably never have noticed.
"I like to lick butts!" by MobileTatsu-NJG (#32700246) (Score:5, Informative)
I think the video game industry will run with this, and make it a great success. However, I think the film industry will play with this (and other motion-capture technologies) for a while, then resort back to good humans. Why? The actors will play along for a while, but their performances will lose authority because audiences will feel their performances were "enhanced" with computer aid. Just like in sports, we want our athletes to perform completely on their own merits, and not with the boost of technology. Sure, digital imagery can make an actor look younger, happier, etc, but ultimately, there will be a "return to purity" movement led by actors and demanded by audiences who are fed up with digital trickery. We want real human emotion, not digitized human emotion.
Health Insurance Quotes