Seeing isn't believing...

[Dynasty]

At this year's Game Developer's Conference, the definitive annual forum for coders, John Carmack, high priest of game graphics, lobbed a grenade at the faithful. Faster graphics engines are no longer it, he says.

So if the engine of game progress is no longer the game engine, what is it? And who, if not John Carmack, is behind the wheel.

Carmack makes the following observation: Computers today are a million times more powerful that when we started out in computing. At this rate, given another decade and a half, scenes that currently take four or five hours to render will be runnning on our monitors at 60fps. In terms of image fidelity, thats pretty much photorealistic.
Carmack's gripe is that game worlds still dont appear much like our own. While graphical quality has improved out of all recognition, it has only shown up various other areas of 'world simulation' in games as woefully feeble.
Outside of a game, for example, a 'cluttered' desk from Max Payne 2 would look like the clean work surface of minimalist architect. A 'crowded' nightclub in Deux Ex 2 would appear to have been struck by anthrax in any real world city.

Addressing this will require the computing power to physically represent and simulate much more 'stuff'. And more importantly, an awful lot of content to be created - and at a cost to still make the profit.
But even harder than creating sufficient content is getting it all to interact sensibly. Carmack admitted that id Software shy away from character interactions, for instance, because its near impossible to make it look right. However you want to define 'look right', we've all been there :E. In comparison with conversations with a real human being, talking a game character is less satisfying - and barely more 'real' - than shouting warnings to Mr Punch.

Solving these kind of problems will require technology to plateau, as only then will artists' tools and what's demanded of then be able to settle down. Just as Hollywood cameramen today dont have to keep learning new models, or how to change the lenses.

What Carmack is saying is that the first big problem of game graphics is almost solved...in a few years, the game's rendering engine will have all the processing power it needs. But rather than sating our appetite, this will only leave us salivating for more from our games, and much of what we'll be baying for can be provided only by new game development production and management techniques, rather that clever hardcore coding.

It might sound a modest warning note, but coming from Carmack it's explosive.

Hmm...sorry about the essay :E but i just had to post this. what exactly is the future of gaming? how else can games be improved?

also soz if you think this is in the wrong forum section...i couldnt think of anywhere else to put it lol.

 

[Maxi]

This reminds me of that IBM guy that said computers will never be used by everyone or something along those lines...

 

[Direwolf]

"I see a world market for maybe five computers."
Although I have to say Carmack has a point.

 

[Synthos]

i don't beleive that's true.
Technology has usually developed at a constant rate (except if we take the quantum leap) and games will always keep up.
In fact, Games such as Doom 3 and Farcry are some of the first to really push hardware to it's limits.
What I predict is that in the next while you'll see many games with doom 3 or farcry quality graphics, while hardware again resumes it's lead over software, then the process will repeat.

but then again, I'm not carmack am I, so what credibility do I have?

 

[MultiVaC]

Of course, games will be become much more work to develop. That's why games like HL2 take longer, cost more and have bigger development teams than older games.

 

[The Mullinator]

There is a limit to how good the graphics can get though, the human eye can only see so much detail. Granted we are still a long way from being photo-realistic it will probably only be at the most 10 years away. After that what happens to graphics? There would be no reason to speed up the graphics card because there simply could not be enough to process to make the image look any better because it can't look any better to us.

One thing I consider that could make the artists job easier would be to have 3d design programs work with completely different concepts than what they currently do. As graphics get more powerful artists have to create more complicated models, more complicated models take more time. Eventually standard 3d modelling as we know it probably won't be feasable and we will have to develop a completely new concept for how to design 3d models, maybe not even use them (computation of particle physics to create objects anyone?).

I didn't get any of that from John Carmack. I have had these views for a long time, Carmack only made them seem more likely to me.

 

[Dynasty]

that's the point. can technology actually get any better than it is? if graphics do become photorealistic, then that will either be their selling point, or their downfall. if graphics are like real life, in the future, what would be the point of games, and what next?

hl2 is a major step forward in the right direction yes, but when you watch the videos as much as i do, you start to notice flaws in design, and you start to realise how this game could look like any other. farcry isnt that far off, if you adjust gamma, brightness etc, then you can make your eyes bleed just by looking at it with a decent g card.

and that's my only argument: where will games go in the future? will they say 'sod the story and gameplay, graphics are what people look for...'

 

[psyno]

Graphics aren't going to plateau any time soon, and the focus should not always to make them photo-realistic. Did anyone here play and enjoy Windwaker on the GC? Anyway there are all sorts of cool things you could add to continue to make graphics more and more realistic (and expensive), like more accurate light models, better, more efficient, or more powerful approaches to anti-aliasing, and of course more and more detail.

If what he's getting at is that the quality of graphics is advancing much faster than the overall quality of games, then I completely agree.