Tim Sweeney has predicted that we'll be having photo-realistic environments in games in another 5 and a half years..
By 2010, I think we'll be photo-realistic in outdoor and indoor environments," claimed Sweeney. "But I think we'll fall very short of reality in the areas of realistic character animation and interaction. A lot of the problems we're facing in upcoming games center around simulating human thought. Right now, when you walk up to a game character in Half-Life 2 and have a conversation, it's pre-recorded... but in the future, it's not going to be."
Sure, it's a long way away (although not TOO long off), but are we excited or what?!
We can get photorealism in 3d rendering programs now, so it's not a surprise to hear that it'll be realtime in the near future, but needless to say, I can't wait. :)
quote:Just do the Doom 3 engine technique...turn off the lights and pretend it has realistic environments.
Wait, you mean the universe ISN'T made out of poorly lit small rooms and corridors?
I'd say his prediction is pretty accurate. We can expect a lot more improvements in graphics realism but I wouldn't expect too much in the way of AI/human thought in only half a decade.
I agree with Aven also: When you're trying to create a photo-realistic scene, the realism is only as effective as its weakest point; the slightest hint of artificiality can blow the illusion.
as much sweeny has the odd debatable point id more or less take his word for it. Considering hes kinda been a part of one of the great leaps into the "photo-real" games. Well visually anyway. Like he mentioned about pre-recorded conversations, theres all these other aspects to make up a realistic enviroment which will probably be even harder then the visual aspect.
Drood: Isn't that amount of money already being pumped into some games? Unless you're talking about the huge blockbusters, well I don't think that'll happen anytime soon, games are just not a big enough market yet. Movies are usually world wide, games aren't there yet.
TheBigJ: Well you could say that the first few were because of technological limitation and the linear nature of other 2d games out at the time, like you could say that wolfenstein was just basically mario but instead of moving up/down/left/right you moved forward/back/left/right... Doom was much of the same, and Quake was that just in 3D. Quake 2 had a story somewhat but it was still linear. But then look again at Half Life, now if that wasn't linear then I don't know what is anymore.
I think the whole corridors thing comes up because its the closest analogy in 3d to the platforms in 2d... But that's just a small though I had just now.
I often talk about what it would be like (and how very cool it would be) to have really high rez detailed models in games and stuff with heaps detailed animation but when I thought about it a bit more I realised hand weighting verts for a 200,000 tri model would be painstaking indeed.
but i hope games get more improvement with content than just making a game because it is 'flashy'.
I find it an odd position when people say developers should focus more on making a game fun then on graphics. It?s not like anyone thinks, ?lets make a purely graphical game but we won?t worry about anything else?! Of course developers want their games to be fun, but I think it?s a good thing to have great graphics also. And when these photo-realistic games come out, I?m sure all the people who said it won?t be that good will be lining up to play them! :P
That said, I think the area that needs the most improvement is game AI. We really could do a lot more then we are doing now.
damien: Most hollywood/american films well exceed even the most funded of games. 200 million compared to 5 million is a pretty huge difference.
Obviously 200 million isn't the norm but they still have more money pumped into them.
Devils advocate for a second: Film's generally have alot more people working on them and a shorter time frame for completion, oh and those pissy actor types who demand ridiculous salaries. [:P]
Auxinamoon: You don't hand wieght something that large. [:P] You generally have a proxy model driving the animations of the larger one.
Well there was that one article that stated if a human face is very lifelike then people will concentrate on the 5% or so that's different or wrong and not believe it, while if the model of a human is not as realistic then they'll focus on the 85% of things that are correct. I forget the article link though...
I guess my point is that if we strive to make things too realistic then the people will focus on the differences and irregularities. Sort of like in "The Matrix" perhaps.
I remember one time watching TV and they had a bunch of Game Developers on an Australian panel discussion show, and one of them (I think from Ratbag) said that he was looking forward to vastly different interactive environments. Sort of like comparing the end worlds of Half Life to the mundane earth environments.
So perhaps it is better to make totally far out worlds with the graphics power that we have, rather than focus on reality too much.
http://slate.msn.com/id/21020
http://slate.msn.com/id/2102086/
Its a purely subjective notion; Some people may feel this way, others won't. It may be difficult for people like us to experience this effect as we've all been playing games for ages and we've been slowly introduced to near-photorealistic characters. I suppose for someone who has previously only watched tv and movies, then moving into games at this time would seem a little strange.
"The only thing that's stopped Sweeney from producing totally photo-realistic games has been the limitations of computer hardware. "Producing realistic simulations of environments is quite simple," he said "We just need enough brute force computing power. We're not quite there yet." "
I don't think that this is the only thing stopping people. The world over, people are trying to create photo real art that isn't real time. Although the power does come into it, it is also a very (artistically) difficult thing to achieve. People are always getting closer, but it is still a way off :)