Future graphics in games - Crytek presentation

[EricMaslowski]

This is great. Thanks for posting this. It's great hearing how industry leads are viewing current trends in graphics technology.

[CGmonkey]

5 years until current quality of offline renders are reached in real-time? Man, that's some bold assertion!

[EricMaslowski]

Yeah, that is bold, but I bet his talking more of "perceptive" quality like what's found in movies now as opposed to accuracy which is often found in ArchViz. I could see real-time graphics looking like a current film rendering in the next 5 years, but I'm not so sure it will be as accurate and able to withstand close inspection like a common VRay architectural rendering...

[CHRiTTeR]

We already can do realtime raytraced reflections and do realtime tesselation/displacement.

If you think about it, we arent 'that' far off to use the same techniques as in movies.
The biggest problem/difference would be the amount of data that needs to be processed to get the same amount of detail as in a hollywood vfx production.

In that regard 5 years may be a little too soon though but if they actually mean using the same techniques then i think they could be and probably are right.

[Krotma]

If a scene takes 5min per frame to render, and Moores law dictates that computer power is doubled every two years... realtime 30FPS animation of that scene will be possible about 25 years from now.

[CHRiTTeR]

moore's law only applies to the evolution of the cpu... it doesnt take the possibilities of new technologies into account or possibly maybe even improvements in coding and instructionsets etc... ;)

also, moore's law is more a raw estimate than anything else.

[Shletten]

Saying that games will reach offline quality in 5 years is bold indeed but feasible! That is as long as it's not as accurate and complex. Some actual games look tremendously good but are limited because of the actual hardware.

I don't know if you guys have played Mirror's Edge but that game looked really good thanks to its pre-baked global illumination! ;)

And we all know that the CryEngine 3 support real-time GI with light propagation volumes. Crysis as a whole looks already damn good and I saw many shaders that look better than what I've seen in some animated shorts.

There's a game named Trico which I think looks really good and cinematic : http://www.youtube.com/watch?v=xF3fED8EXl4 Huh, it's now called The Last Guardian. http://www.gametrailers.com/video/trailer-comparison-the-last/51781

What we most need right now are improvements in lighting, texture resolution and animation. Geometry is getting at a point where it's almost alright and tesselation is going to help for sure.

[Kabab]

Some games/engines can already do similar quality to Toy Story 1 now so I don't see how the claim is so bold...

[Laa-Yosh]

Games have seriously compromised image quality and other limitations.

Antialiasing, motion blur/dof, texture sharpness (and resolution!), shadow quality, GI precision and quality (Crytek's solution only works in very, very small, confined spaces) and so on are all way behind an offline renderer.
Or what about the stuff that is incredibly difficult to do interactively, as we can only get by using cheats? Dynamics simulations with per frame fixes for cloth, hair, muscles, collisions, deformations... how will a game, where no TD can come to the rescue, reach such levels of precision?

And what kind of quality are we talking about anyway? The game intro cinematics that a few studios push out with a few months' worth of production time? Or something like Transformers 2, with multi-million polygon robots and raytraced reflections and lighting that took days to render on ILM's farm? Or Avatar with its SH lighting and super detailed characters?

Games will look better, no doubt about that, but let's not get unrealistic here.

[705]

i personally think the future gaming is in cloud gaming. imagine a computer having 50GB of memory and everyone sharing it. i think its more efficient.

[simonenastasi]

That's actually being developed, don't remember the leading company's name. But I don't see how the lag problem could be solved in 5 years...

[CHRiTTeR]

That's actually being developed, don't remember the leading company's name. But I don't see how the lag problem could be solved in 5 years...


And thats what they (the big companies) are all trying to push...

I seriously hope this is not going to happen. I like the idea of having MY own data on MY OWN pc and not having to relly on some remote servers of which i have verry limited control, not to mention personal documents etc...


If its for rendering/computing allone, i guess its ok... but i doubt it will be like that.

[BigPixolin]

I think it is called onlive, and I too hope that gaming doesn't go this way.

[Shletten]

Games will look better, no doubt about that, but let's not get unrealistic here.
I agree but I for one am not asking for real-time fluid simulations in 5 years. :surprised

[CGTalk Moderation]

This thread has been automatically closed as it remained inactive for 12 months. If you wish to continue the discussion, please create a new thread in the appropriate forum.