Epic's Tim Sweeney reckons that not only are photorealistic games inevitable, but will be with us soon.
Epic's Tim Sweeney has spoken out on where he thinks the graphics industry is heading, claiming that he thinks truly photorealistic games are not only inevitable, but will be upon us within the next decade.
"
We're only about a factor of a thousand off from achieving all that in real-time without sacrifices," Sweeney said in an interview with
Gamasutra. "
So we'll certainly see that happen in our lifetimes; it's just a result of Moore's Law. Probably 10-15 years for that stuff, which isn't far at all."
Sweeney clarifies what he means by photorealism too, just to leave no shadow of a doubt. Photorealism consists of totally realistic lightning with real-time radiosity, perfect anti-aliasing and movie-quality animations and static scenes. He's confident that the issue can be solved by a brute force approach, so the only issue is generating the computing power to manage it. On that front ten years isn't such a long time either if you compare games from 1999 to games from this year; say,
System Shock 2 to
BioShock.
The real challenge as far as Sweeney can imagine will be AI and emulating the nuances of human behaviour without falling into the uncanny valley.
"
A state-of-the-art game like the latest Half-Life expansion from Valve, Gears of War, or Bungie's stuff is extraordinarily unrealistic compared to a human actor in a human movie, just because of the really fine nuances of human behavior," he said.
"
We simulate character facial animation using tens of bones and facial controls, but in the body, you have thousands. It turns out we've evolved to recognise those things with extraordinary detail, so we're far short of being able to simulate that."
Let us know your thoughts on where the games industry is heading in
the forums.
1993:
http://upload.wikimedia.org/wikipedia/en/d/de/Doom_ingame_1.png
1996:
http://mac.softpedia.com/screenshots/9-548_1.png
1999:
http://unrealitymag.com/wp-content/uploads/2009/01/quake-3-arena.jpg
2002:
http://upload.wikimedia.org/wikipedia/en/thumb/a/a0/Morrowind1.jpg/800px-Morrowind1.jpg
2005:
http://www.counterfrag.com/screenshots/fear%20game/1.jpg
2008:
http://www.gameguru.in/images/crysis-warhead-ss2.jpg
To be honest, since 2002, games have looked pretty damn good, and ever since then all we've really managed to do is increase the pretty post-processing effects and increasing texture size. We're still churning polygons in the same way Quake did all those years ago. I mean, given the best looking game out right now came out a year and a half ago, it's not exactly like graphics are thundering along at a blistering pace now are they? I think the hay-day of every game breaking new graphical barriers are coming to an end, and we'll see a real slow down in the improvements people are making to game engines year on year.
Agree and disagree. It's true with any line of improvement, ultimately there are limited returns. But just as you say post-processing has got better - surely anything looks better with post-processing? So I think graphics is definitely moving forward, that doesn't have to just through continuous improvement of FPS and such but also as the industry evolves and grows there is much more ingenuity and new ideas. For example, procedurally generated textures may be the way forward but they're not implemented in any games yet - it's all still down to the artist.
Look here: http://www.bit-tech.net/gaming/2006/11/09/Procedural_Textures_Future_Gam/1
http://images.bit-tech.net/content_images/2006/11/Procedural_Textures_Future_Gam/01.jpg
And without a doubt there will be new ideas and concepts (the majority of which probably wont make it through to the end project) that will continue to open new boundaries and new possibilities that aren't just a linear progression of today's tech.
To be honest though, if they just made every game look as good as Crysis with some of the texture mods etc, then I would be happy. I'm pretty sure if everyone could play a modded Crysis maxed out with 8x AA and all that jazz, no one would be complaining. But still, I do love progress!
Not necessarily animation and physics, but better textures, models, shaders and other effects, multicore support etc.
Never gonna happen commercially though, there's only fans doing that sorta work :(
Its AI, physics and great storylines that are most needed in new games. Oh, and the bottom to fall out of the console market.
I'm sure there are still some gob-smacking revelations to come- but graphics are being overdeveloped while AI, animation, design and quality control can barely squeeze the time and/or money they need. Priorities please!
QFT.
Games don't always have to look good, take System Shock for example, the graphics are horrible by todays standards, but the gameplay still is among the best of any game.
STALKER: CS, not the best game, but it's atmosphere, story, and just general aura is much better than Crysis, hence why I'd enjoy playing that rather than Crysis.
In other words, as delriogw said, quality over looks and quantity.
I think in the future games will be just as powerful as a good movie.
Making you laugh, cry, fear, guilt and even anger when a child in a hostage situation for example gets killed.
You don't really care at the moment because the graphics are still crude, hell, people might even end up falling in love with characters.
Real emotion while your playing that the crude graphics of today can't really manage.
In 15 years time crysis will look as crude as doom does now, and we'll all laugh about how we all thought crysis in the day looked great.
I say bring it on, i would love a gaming experience like that.
Our survey says.... ERRRR!
Graphics have nothing to do with emotional connection. Ever read a book? Doesn't even have any graphics, but a good novel packs 100 times the emotional content of something like Cyrsis (hell, make that 1,000 times).
It's the script, writing and delivery that makes a game emotional. Graphics are totally and utterly irrelevent.