Unreal Engine Plans

A post to the BeyondUnreal Forums (thanks HomeLAN Fed) has word from Epic's Tim Sweeney on where they've been focusing their efforts for the next iteration of their Unreal engine. Here's a bit:
We already have polygon count where it needs to be. We're going to be focusing on per-pixel rendering and getting the quality of each pixel on screen as high as possible. So there's a bunch of different technologies there: dynamic shadowing technology, stencil buffer and shadow z-buffers. But the whole idea is that every object in the scene should cast realistic shadows with respect to every light source in the scene, and every pixel you see should illuminate properly that way. Of course that's easy to say and there are some solutions for that, like stencil buffering. But to do that properly you really want fuzzy shadows everywhere because most light and most environments are quite diffused and if you look around you very seldom see a really sharp shadow edge somewhere. So a huge amount of effort and processing power goes into implementing fuzzy shadows effectively in real-time. That's been a significant part of our R&D right there.
View : : :
32 Replies. 2 pages. Viewing page 1.
Older [  1  2  ] Newer
1.
 
Yeah but
Jan 25, 2004, 12:52
1.
Yeah but Jan 25, 2004, 12:52
Jan 25, 2004, 12:52
 
Can the engine do water? Big levels? I'm still smarting from DXIW

2.
 
Re: Yeah but
Jan 25, 2004, 13:03
2.
Re: Yeah but Jan 25, 2004, 13:03
Jan 25, 2004, 13:03
 
Ah, I likes me some fuzzy shadows. DXIW could've just saved themselves the trouble. Great big sharp shadows look so bad.

3.
 
Yeah, it's pretty, but...
Jan 25, 2004, 13:07
3.
Yeah, it's pretty, but... Jan 25, 2004, 13:07
Jan 25, 2004, 13:07
 
I like games with nice graphics. But, it's the gameplay that should be the focus (for all game developers) seeing as that's why (or should be why) we actually p-l-a-y games...

Sometimes, I think too much has emphasis has been put into how good a game looks rather than how good it is to play. Look at the majority of the top selling games on the market from month to month/year to year. They are usually less than the highest quality graphics. Why? Because most ppl purchasing games don't have the hardware to run those "prettier" games well.

Yep, it's the "average Joe" gamer that drives the sales market. Eh, that's probably why so many game companies (developers/publishers) focus on making the box art and screenshots look better than some of the games...lol.

4.
 
Re: Yeah, it's pretty, but...
Jan 25, 2004, 13:22
4.
Re: Yeah, it's pretty, but... Jan 25, 2004, 13:22
Jan 25, 2004, 13:22
 
I agree with you, but that wasn't the point of this post. It was just a heads up as to where the Unreal Engine is going into the future... so the gaming community knows what to expect, this really has nothing to do with gameplay. And one thing that's nice about the Unreal Engine is they keep it extremely scaleable. I can still play UT2003 on my Athlon 700 / GF2 Ti machine, it's not nearly as pretty as on my updated machine, but with some settings tweaks I can get it up to 40 FPS.

Avatar 13977
5.
 
Re: Yeah, it's pretty, but...
Jan 25, 2004, 13:32
5.
Re: Yeah, it's pretty, but... Jan 25, 2004, 13:32
Jan 25, 2004, 13:32
 
I agree with you, but that wasn't the point of this post. It was just a heads up as to where the Unreal Engine is going into the future... so the gaming community knows what to expect, this really has nothing to do with gameplay. And one thing that's nice about the Unreal Engine is they keep it extremely scaleable. I can still play UT2003 on my Athlon 700 / GF2 Ti machine, it's not nearly as pretty as on my updated machine, but with some settings tweaks I can get it up to 40 FPS.

Agreed. Its pretty safe to assume that the days of "UNREAL ENGINE SUX!!" are over.

can't wait to see some screenshots of the new tech...(and maybe a hint at what "Unreal Warfare " is all about)

----------------------------------------------------------------------
"Perhaps the answer to the perennial problem of delinquent teenagers dropping bricks from motorway and railway bridges is to sue the creators of Tetris."- unknown author
----------------------------------------------------------------------
"Both the “left” and the “right” pretend they have the answer, but they are mere flippers on the same thalidomide baby, and the truth is that neither side has a clue."

- Jim Goad
Avatar 10137
6.
 
Re: Yeah, it's pretty, but...
Jan 25, 2004, 13:34
6.
Re: Yeah, it's pretty, but... Jan 25, 2004, 13:34
Jan 25, 2004, 13:34
 
this really has nothing to do with gameplay.

Unfortunately too often it does. Tiny maps with lots of transitions have a big effect on gameplay. Goodbye sniper rifle. Who needs it if everything is within a few yards of you. Goodbye AI and monsters following you if all you have to do is lose them at a new map. If all the new graphical bells and whistles means theres no more CPU/GPU left for water, then no swimming, no aquatic monsters, etc.

I suppose what the developer does with the technology is an important consideration, but one must also consdier what they were given to work with too.

7.
 
Re: Yeah but
Jan 25, 2004, 13:39
Quboid
 
7.
Re: Yeah but Jan 25, 2004, 13:39
Jan 25, 2004, 13:39
 Quboid
 
Water is what I'm looking for in future engines. Current game poly-based technology is pretty good for solids but not for liquid or gas. Gases are made up of sheets of transparent polys, while liquid is made from a pixel/vertex shaded polygon. Neither of those are very realistic but I've no idea how to improve this without the huge processing power required to do stuff like modeling the material as a 3D matrix of points at as high a resolution as possible.

Asside from that, lighting is where people seem to be concentrating and that makes sense. Poly count is now high enough so art quality is more important than the number of triangles, while the ~700 polygon Quake 1 models and their low-res textures would waste high detail artwork (not saying Adrian Carmack and the guys did a bad job). Games do need fuzzed shadows - the ones in 3dMark 2003's tomb raider ripoff are really quite nasty for most situations while the Doom3 alpha's looked like solid objects floating around (or so I've heard ), although no doubt these will improve.

Quboid

Currently playing:
CM0304, CoD, FS2004, SC4:RH, Trackmania, X2
Avatar 10439
8.
 
No subject
Jan 25, 2004, 13:40
8.
No subject Jan 25, 2004, 13:40
Jan 25, 2004, 13:40
 
screw water

9.
 
Re: No subject
Jan 25, 2004, 14:07
9.
Re: No subject Jan 25, 2004, 14:07
Jan 25, 2004, 14:07
 
I'd like to see you try it. Post a vid for us ok?

Dr. Michael Bolton's Hairdreser PhD, MD, LLP, DDS
10.
 
Hmmm
Jan 25, 2004, 14:14
10.
Hmmm Jan 25, 2004, 14:14
Jan 25, 2004, 14:14
 
Wast that a quote from the Ion storm team that is working on thief 3?

11.
 
Re: No subject
Jan 25, 2004, 14:15
11.
Re: No subject Jan 25, 2004, 14:15
Jan 25, 2004, 14:15
 
The Unreal engine as it stands right now is damn good. Unreal 2, UT2k3, and other recent releases in the series look and run nicely. Water, reflective surfaces, and even the shadows are very impressive. The problem is all other games that use the technology (Invisible War comes to mind) gut the engine and somehow makes it suck in every way possible.

Avatar 15920
12.
 
Re: Yeah but
Jan 25, 2004, 14:32
12.
Re: Yeah but Jan 25, 2004, 14:32
Jan 25, 2004, 14:32
 
"Poly count is now high enough so art quality is more important than the number of triangles,"

Generally I know what you mean, but I don't think this is the case overall. It all depends on what type of game you are making. If you make something with tight indoors then maybe, poly counts are almost ok. But go outdoors and there is a frightening lack of detail in most outdoor, real time environments.

13.
 
Water
Jan 25, 2004, 14:38
13.
Water Jan 25, 2004, 14:38
Jan 25, 2004, 14:38
 
while liquid is made from a pixel/vertex shaded polygon. Neither of those are very realistic

I dunno ... I've never seen anything as amazing as the water in Morrowind. It looks pretty damn realistic to me. From the videos the water in HL2 looks to be pretty nice too. I can't imagine what you could do to make it better. Waves and foam would be nice. A really weird omission in Morrowind considering how large those moons are.

14.
 
Re: Water
Jan 25, 2004, 15:05
14.
Re: Water Jan 25, 2004, 15:05
Jan 25, 2004, 15:05
 
Check out the Far Cry demo for a great example of how water can look in a game.

Alex
-----------
Spatial Fear
Director/Creative Designer
http://www.planetunreal.com/le/sf/
15.
 
Re: Water
Jan 25, 2004, 15:38
15.
Re: Water Jan 25, 2004, 15:38
Jan 25, 2004, 15:38
 
And to see what "dynamic stenciling with soft filters" does for shadows.

If you crank up all the effects, the individual leaves on the foliage cast shadows on your gun in FarCry...and they don't have sharp edges...they are diffuse.

-^D^

16.
 
Re: Water
Jan 25, 2004, 15:59
16.
Re: Water Jan 25, 2004, 15:59
Jan 25, 2004, 15:59
 
Yes, Far Cry's water is impressive but there is still room for improvement. I think water is going to be the first detailed effect that will be photo realistic in the coming years, they are almost there now.

"When all thats left is console gaming, I will game no more."
Avatar 16493
17.
 
No subject
Jan 25, 2004, 15:59
17.
No subject Jan 25, 2004, 15:59
Jan 25, 2004, 15:59
 
Waves and foam can be tough to do well, but there was a good paper on Gamasutra on it: http://www.gamasutra.com/gdce/2001/jensen/jensen_01.htm (that'd be registration required).

The Unreal engine as a whole isn't incapable of doing fairly large outdoor areas - examples can be seen in U2-XMP. The unreal engine is pretty damn flexible and powerful, so this really isn't about them failing to focus on providing developers with the engine tools to produce great gameplay. Blame the individual developers for their own design choices.

There's also nothing wrong with taking advantage of pixel and vertex shaders in water; if you particularly think they're the things holding everyone back from producing the next and best wet stuff, I'd say you're mistaken. Take a look at some of the Cg (the NVidia PS/VS language) demos that people have done to see what sort of power Pixel and Vertex shaders have. The current technology that's the one up of using polygons is volumetric particles, but frankly, it's not worth it. Grab a high end 3d workstation and render good smoke with those and you're usually talking several minutes and upwards of render time, they're simply not very practical for the graphical payoff in realtime games. Smoke and gas can be fairly well faked using polygons at a fraction of the cost.

I'm still waiting to see something that makes good use of Unreal facial (?) and lipsyncing tech which they've mentioned in the past. That would certainly be one of the things that I'd like to see really well done by the engine in future. The faces in the Deus Ex:IW demo (still not out here) *really* sucked, I've seen better looking faces in much older games than that of Alex Denton in that, so it's certainly something I'm becoming more demanding about.

18.
 
I don't know...
Jan 25, 2004, 16:10
18.
I don't know... Jan 25, 2004, 16:10
Jan 25, 2004, 16:10
 
I mean, it sounds very impressive and such, but is it really what we want, or even NEED, in our next generation engines? The Unreal engine is predominantly used for shooters, why do I need per pixel rendering of shadows? I'm not looking at my enemies' shadows, I'm looking at them, and where I'm shooting them.

We can do pretty convincing shadows in current engines already as well, and usually I just turn them to low so that the shadow looks like a big blob on the ground. More than enough for me. I'm not sure if devoting that much processing time to strictly shadows is what we're really looking for...

Creston


Avatar 15604
19.
 
Re: No subject
Jan 25, 2004, 16:17
Quboid
 
19.
Re: No subject Jan 25, 2004, 16:17
Jan 25, 2004, 16:17
 Quboid
 
Farcry and Morrowind (and PotC and DF:Black Hawk Down) certainly have water that looks very nice, and look very realistic, however they're not so great when you're in. They could improve that by simply having better effects but even if it looked right, it wouldn't be at all realistic physically. Ripples and waves shouldn't be artwork drawn on the surface, although sometimes this looks excellent (check out Farcry's shores - no nasty line where the water ends, but nice rippling wavelets) and water should move if there is anywhere gravity would pull it into. Likewise with gas - things like clouds in flight sims forming and floating realistically and murky dungeons having mist floating around (which I think would look unbelievably cool - think dry ice from nasty 80's music videos, the way it flows and swirls as things move through it and has it drifts off a ledge - that in a Morrowind dungeon would look awesome) would be nice, sometimes.

I guess I just don't like games trying to be realistic in such a fundamentally unrealistic way. Someday I hope this improves, but I know it won't be a while - polygon tricks are just so much more efficient than point modeling so polygons will have to do until the power becomes available. I believe it will, just like polygons are less efficient than sprites but when the power became available, the flexibility and visual effects took priority.

Quboid

Currently playing:
CM0304, CoD, FS2004, SC4:RH, Trackmania, X2
Avatar 10439
20.
 
Re: No subject
Jan 25, 2004, 16:29
20.
Re: No subject Jan 25, 2004, 16:29
Jan 25, 2004, 16:29
 
"The Unreal engine as a whole isn't incapable of doing fairly large outdoor areas - examples can be seen in U2-XMP."

Some of the UT2004 maps were enormous to the point that they had to reduce their size because their size was detrimental to gameplay. XMP != 2004. I assure you all, after playing UT2004 at a lan party at an Epic sponsored event that UT2004 delivers in its gameplay, at least in the onslaught mode(didn't get to play assault).

This comment was edited on Jan 25, 16:30.
32 Replies. 2 pages. Viewing page 1.
Older [  1  2  ] Newer