Future game engines?
Future game engines?
Correct me if I'm wrong, just thinking aloud. It seems id software have been responsible for the major game engines over the years.
Quake3 engine has been licensed to quite a few games and now the Doom3 engine, including the upcoming Prey. I don't even think valves own engine is THAT far removed from the D3 engine. Certainly it's a fantastic progression from it but I don't see it as revolutionary as say q3 to doom3 /Q4, at least, in looks.
So do you think the next major games engine will come from id again and when? Another 3 or 4 years?
Quake3 engine has been licensed to quite a few games and now the Doom3 engine, including the upcoming Prey. I don't even think valves own engine is THAT far removed from the D3 engine. Certainly it's a fantastic progression from it but I don't see it as revolutionary as say q3 to doom3 /Q4, at least, in looks.
So do you think the next major games engine will come from id again and when? Another 3 or 4 years?
The differences between Q3 and D3 engines dont seem that revolutionary to me. Bump mapping and fully dynamic lighting pretty much top the billing and they're just not that impressive.
An extensively modelled physics engine which allows you to build a wide randge of contraptions from basic elements, on the other hand, is very impressive.
I'd be more supportive of the latest id engines along the way BUT we seem to be losing features and not gaining them. Right, so you've strapped on per-poly hit detection, but now I can only really hold a 4 player deathmatch, and a clunky one at that. Oh, you've added fully dynamic lighting, but you can only add about 5 of them to a room before the game crawls to a halt, and you REMOVED the option for static lightmaps as well.
Anyways, the UT engine is gaining popularity too. It was already pretty popular even back with the first version of the engine. I think it was used for more games than Q3's engine although I don't have the time to find sources for that. Moving forwards, UT2003/4's engine provided a steady move forwards. Not millions of extra bells and whistles but the technologies it did implement were used to their fullest extents to create some really beautiful scenes and a very stable basis for actually getting on and playing the game.
I think relatively few companies are licensing the D3 engine... it's simply not very flexible and performance is weak given the level of graphics it can push. It doesn't even scale well. unless id's next engine, if they even do one, is built using a completely different paradigm to whatever goes on within id currently, the increasing gap between what id and other vendors are outputting is only going to get wider.
An extensively modelled physics engine which allows you to build a wide randge of contraptions from basic elements, on the other hand, is very impressive.
I'd be more supportive of the latest id engines along the way BUT we seem to be losing features and not gaining them. Right, so you've strapped on per-poly hit detection, but now I can only really hold a 4 player deathmatch, and a clunky one at that. Oh, you've added fully dynamic lighting, but you can only add about 5 of them to a room before the game crawls to a halt, and you REMOVED the option for static lightmaps as well.
Anyways, the UT engine is gaining popularity too. It was already pretty popular even back with the first version of the engine. I think it was used for more games than Q3's engine although I don't have the time to find sources for that. Moving forwards, UT2003/4's engine provided a steady move forwards. Not millions of extra bells and whistles but the technologies it did implement were used to their fullest extents to create some really beautiful scenes and a very stable basis for actually getting on and playing the game.
I think relatively few companies are licensing the D3 engine... it's simply not very flexible and performance is weak given the level of graphics it can push. It doesn't even scale well. unless id's next engine, if they even do one, is built using a completely different paradigm to whatever goes on within id currently, the increasing gap between what id and other vendors are outputting is only going to get wider.
Last edited by Foo on Thu Mar 02, 2006 4:03 pm, edited 1 time in total.
Well don't forget Epic. Their Unreal Engine 3 is already far more popular (more licenses) than the Doom3 engine. In fact, most of the Doom3 engine licensees so far that I know of are long time id Software "friends". Raven (Quake 4) and Splash Damage (QW ET). Other than that there's Humanhead studios developing Prey.
a game engine that does graphics ala this modeling app; http://pixologic.com/home/home.shtml
thats what the future is at
and I think neither ID nor EPIC, nor any 'widely' known game dev will develop such an engine. then again, who the fuck am I ;O
thats what the future is at
and I think neither ID nor EPIC, nor any 'widely' known game dev will develop such an engine. then again, who the fuck am I ;O
Last edited by corsair on Thu Mar 02, 2006 4:46 pm, edited 1 time in total.
Modelling apps can't churn out 60FPS on widely available hardware. Realtime rendering will always be steps behind prerendered scenes.
Prerendered stuff from 5 years ago look poor in comparison to modern realtime rendered scenes. In 5 years time I would expect to see realtime stuff comparable to what you're showing there.
Of course, there's still the looming problem of the amount of development time needed for realtime content - it's increasing exponentially; id software seems like a big casualty of this - they never expanded to a size capable of delivering the next generation engine and game. You can see Epic and Valve expanding to meet this need.
Prerendered stuff from 5 years ago look poor in comparison to modern realtime rendered scenes. In 5 years time I would expect to see realtime stuff comparable to what you're showing there.
Of course, there's still the looming problem of the amount of development time needed for realtime content - it's increasing exponentially; id software seems like a big casualty of this - they never expanded to a size capable of delivering the next generation engine and game. You can see Epic and Valve expanding to meet this need.
-
- Posts: 1328
- Joined: Sat Nov 05, 2005 1:00 am
doesn't half-life use havoc for the physics?
havoc does some extremely cool things. quake wars looks like being the first game that might show off what the d3 engine can do in comparison, but i can't imagine it being quite as feature-packed as the havoc stuff, because havoc is a dedicated physics thingy (that's a technical term) that valve have bought in rather than developing themselves.
havoc does some extremely cool things. quake wars looks like being the first game that might show off what the d3 engine can do in comparison, but i can't imagine it being quite as feature-packed as the havoc stuff, because havoc is a dedicated physics thingy (that's a technical term) that valve have bought in rather than developing themselves.
-
- Posts: 357
- Joined: Tue Nov 27, 2001 8:00 am
The rocks and the rest of the background are real. The only thing that's "fake" in those pictures are certain elements of the creature.DRuM wrote:Holy shit, those images from corsair are insane. I thought they were photos, especially the one with the lorry driving away. And the rocks on the floor look so real
Yeah, I think I've read somewhere that Valve did a bunch of in-house modifications to Havoc though.4days wrote:doesn't half-life use havoc for the physics?
havoc does some extremely cool things. quake wars looks like being the first game that might show off what the d3 engine can do in comparison, but i can't imagine it being quite as feature-packed as the havoc stuff, because havoc is a dedicated physics thingy (that's a technical term) that valve have bought in rather than developing themselves.
Also, been hearing some buzz-words thrown around that DNF will have physics more advanced than HL2 (Macon it's called or something, also done by a 3rd party, but prolly customized by 3drealms for their needs).
It might also take advantage of one of those hardware based physics cards, but imo that's just crap, add yet another hardware piece for a marginal performance gain.

I prefer to forget Epic. Having purchased Unreal, Unreal Gold, UT, UT2003, UT2004, and various UEngine games, I have seen no indication that Epic or Atari plans to release the UT99 source code. That sucks, since I truly loved the Unreal experience and have fond memories of UT circa '99-2001. Unfortunately, I won't be able to relive those memories outside of emulation in the future.Eraser wrote:Well don't forget Epic.
No, I'm not bitter. Really.

-
- Posts: 448
- Joined: Tue Mar 19, 2002 8:00 am
Source is just a glorified Q1 branch, and Carmack isn't all too worried about it as id still gets the royalties from source profits from still licensing Q1 tech.
Leader and director of the [url=http://cheapy.deathmask.net]OpenArena[/url] project which is a free software version of q3a designed for hobo fagts
-
- Posts: 4022
- Joined: Sat Mar 12, 2005 6:24 pm
Not that marginal. Remember the difference between Q2's software and (hardware accelerated) OpenGL renderer? Realistic physics are complex and expensive, performance-wise, so the more you can offload to external processors, the smoother and better looking your gaming experience will be.dzjepp wrote:It might also take advantage of one of those hardware based physics cards, but imo that's just crap, add yet another hardware piece for a marginal performance gain.
-
- Posts: 1328
- Joined: Sat Nov 05, 2005 1:00 am
Havoc isn't the most advanced physics engine anymore UE2 uses havoc, UE3 will use something completely different (can't exactly remember what it was called). As for hardware physics, the only thing I can think of how that's going to work is if it's integrated into videocards so you don't need to buy an additional card.dzjepp wrote:Yeah, I think I've read somewhere that Valve did a bunch of in-house modifications to Havoc though.4days wrote:doesn't half-life use havoc for the physics?
havoc does some extremely cool things. quake wars looks like being the first game that might show off what the d3 engine can do in comparison, but i can't imagine it being quite as feature-packed as the havoc stuff, because havoc is a dedicated physics thingy (that's a technical term) that valve have bought in rather than developing themselves.
Also, been hearing some buzz-words thrown around that DNF will have physics more advanced than HL2 (Macon it's called or something, also done by a 3rd party, but prolly customized by 3drealms for their needs).
It might also take advantage of one of those hardware based physics cards, but imo that's just crap, add yet another hardware piece for a marginal performance gain.
While I agree that the limitations of the D3 engine are almost throwing us a couple of years back in time, but you shouldn't underestimate the power of real time dynamic lighting. I wouldn't take D3 as a benchmark for the D3 engine either. The ETQW screenshots already look far more impressive than any D3 screenshot (or Quake 4 screenshot for that matter) ever released. I hope ETQW will clear up a lot of the limitation issues D3 and Q4 have.Foo wrote:The differences between Q3 and D3 engines dont seem that revolutionary to me. Bump mapping and fully dynamic lighting pretty much top the billing and they're just not that impressive.
An extensively modelled physics engine which allows you to build a wide randge of contraptions from basic elements, on the other hand, is very impressive.
While physics are absolutely fantastic when used in an effective way, they're not nescessarily the definition of a great engine. I didn't really miss an advanced physics engine in Doom3 or Quake4. In fact, I thought HL2's physics were way overrated (I actually enjoyed them more in Max Payne 2 than HL2)
I do agree though that id Software are losing focus and are indeed losing the battle of the #1 innovator spot quickly. Epic is gaining in on them. However, I'll wait with judging them until I see what id Software comes with by the time the first UE3 game appears.
History has shown, and I believe this will be the case again, that Carmack will absolutely blow us away with the next engine.
They are pretty much always the guys who usher in the new generation, and I think the follow up to the DOOM 3 engine is going to be significantly stronger than Unreal Engine 3. But, we may not see it on the market for a couple of years.
They are pretty much always the guys who usher in the new generation, and I think the follow up to the DOOM 3 engine is going to be significantly stronger than Unreal Engine 3. But, we may not see it on the market for a couple of years.
This line only remake is total rubbish I've ever seen!!! Fuck off!!! --CZghost
I agree Mogul. However, it seemed to be a good 5 years between the Q3 engine ( 1999) and the D3 engine ( 2004) so maybe it would still take another 3 and a half to 4 years, since D3 came out over a year ago if there is any sort of similar pattern to their development process. Unless technology is progressing at a faster pace these days?
Carmack's aiming for more gradual improvements to existing engines rather than a new engine iteration every 4 or 5 years. He said that after the D3 engine they would probably do one major engine revision to make use of the current generation hardware (D3 engine uses GeForce 3 generation hardware as a feature set guideline).
After that they'll probably reuse that engine and add individual features. This can allready be seen in the D3 engine with stuff like MegaTexture technology in ETQW.
After that they'll probably reuse that engine and add individual features. This can allready be seen in the D3 engine with stuff like MegaTexture technology in ETQW.