Game Development Community

TSE Scene/Terrain Questions

by Patrick Bohnet · in Torque Game Engine Advanced · 11/11/2004 (3:08 pm) · 34 replies

I have some questions about the TSE. I have read though the forums and what information I can, so hopefully these won't be to many duplications.

The terrain engine in TSE is supossed to be huge when compared to TGE. How flexible is it. I know, with some major tinkering with code, you can make the TGE scene manager seamless, what about TSE... does it come out of the box as seamless, or would that also require tinkering? OR is it so huge that making it seamless would not matter?

Can maps be generated on the fly? I would assume no, unless the "seamless" feature worked, and you could then generate a map on the fly by simply picking the different parts and stichign them togethor. (think of a deungon crawl... where each room is pre built... but how they connect togethor is randomized)

Can the map be deformed during play? lets say a holy hand grenade goes off, and I want to symulate a huge crator? I could think that this might be done... if the map was actually flat and what the users saw as the map was actually all entities...

And this is totally for me... how well would the new TSE "landscape" engine handle a dense cityscape? think venice dense, but with skyscrapers? I know that TGE was originaly designed for outdoors... but would this still count as outdoors?
Page«First 1 2 Next»
#21
12/06/2004 (1:10 pm)
Well, its all discussion really.. If anyone minds, speak up. ;)

But ya, hl2 is more advanced art wise however doom 3 has more power. So really when ya think about it, if you were t othrow in hl2 content in doom 3... should look very stunning and awesome indeed. ;)
#22
12/06/2004 (4:38 pm)
If, that is, the D3 engine could accomodate the extra use of texturing on top of all that shader & normalmapping and not choke its host cpu. HL2 proves there is much life left in other techniques, & that lower platforms can even be accomodated by focusing on the art more than just the science.

i return to the theory that HL2 succeeded so brilliantly with lower cost by focusing on the quality & realism of textures, using photographic textures, and not relying so heavily on shaders to do all that work. the human perception is hardwired to distinguish between the grit of reality and the tinfoil taste of artiface alone. a mix is always best, but the option shld be there. most may think this too labor intensive, but there'll always be those of us who feel the extra 'hand-crafted' effort is worth it, & imho any next-gen engine shld be able to accomodate such different approaches, allowing for work done with hand-wrought/photo textures to succeed in lieu of shading, if a developer wishes that.

[sorry for warping the thread; it's solely out of concern that as much forethought & flexibility as possible goes into the engine, and i started from experience with terrains. i'm glad the point is made & discussion is going on, but i promise not to belabor it from here on out.] -aq
#23
12/07/2004 (7:32 pm)
I am new at working in the Torque Engine and am curious as to it's limitations. What is the max terrain area you can have before it repeats. In other words, if I wanted to make a very large map without any repition, can it be done? And, if it can be done, how do I set the area?
#24
12/08/2004 (10:56 am)
HL2 has normalmaps on a lot of their interiors. Don't forget that - it's not all fixed function. TSE took the same route - combine lightmaps with normalmaps instead of going all realtime like D3. I think it's more flexible for level design and it currenly performs better.

HL2's realism came from the photo driven textures as much and maybe more than the lighting though. It would be interesting to see a D3 based game that was using photo textures.
#25
12/08/2004 (11:59 am)
Hmm.. you said lightmaps and normalmaps for TSE.. Does this mean there are no plans on GG's part to make realtime world shadows like in doom 3?
#26
12/09/2004 (10:02 am)
Brian; i myself have extreme interest in photorealism for texturing, so i will definitely be doing a lot of this. i also hope to start using ZBrush for part of my texturing workflow to more tightly integrate photographic textures into the sculpting of the model itself, and then outputting normalmaps from that.

so even if i don't have a full on game, i will certainly post here as i have walk-through-able experiments in this kind of texturing up and ready. i'm very glad to hear that this approach is being watched and is a guiding force for the team.
#27
12/09/2004 (4:57 pm)
Jacob - there will be realtime shadows, but not D3 style - at least not on TSE release 1.0. There's just not time for it. John Kabus has shown some interest in possibly doing that for a TSE lighting pack though.

D3 dynamic shadows are not a great general case solution anyway. If you have lots of lights casting shadows, your fillrate goes through the roof. Shadowmapping solves the problem (somewhat) on fillrate, but doesn't send light in all directions very effectively like shadow volumes do. Both cases just really suck, but shadowmapping is getting better and there will be better hw support for it in the future.
#28
12/10/2004 (1:32 am)
Hm.. Thas a shame, i was really hopin to able to get a look like doom 3 goin.. O well, maybe eventually.
#29
12/10/2004 (4:14 am)
Well that's the great thing about TSE, you can get it to do whatever you want, you have the source, you'll just need to invest a little effort to differentiate your game from the rest and make it have the look you want.
#30
12/10/2004 (3:59 pm)
Exactly right. The lighting really depends on what kind of game you want to do. Most games probably don't want to burn most of their rendering time on shadows. If that's going to work for you though, then go for it! Doom 3 looks fantastic.
#31
12/10/2004 (6:01 pm)
Heh.. ya, only problem is for fairly noob coders like ymself, saying "put it in yourself" is easier said then done.. :P
#32
12/10/2004 (8:08 pm)
Although i have not yet seen D3 live, i expect that with the TSE in place the Lighting Pack will be able to produce some exceppent effects... there is no single silver bullet in a good looking game, i find, but rather a mix of well used techniques...
#33
12/11/2004 (7:31 am)
Ya, hopefully.. I was countin on full dynaimc shadows to help make a scary atmosphere.
#34
12/11/2004 (1:37 pm)
You know, you could always count on subtle and well-done sounds and disturbingly textured objects and an environment whose architecture whispers 'uneasy' to help make a scary atmosphere as well. that way the shadows won't be pulling so much of the design load all by their lonesome. ;) :) 'atomsphere' is definitely an admirable goal, and definitely relies on a combo of techniques all working in common.. anyway, we shall see!
Page«First 1 2 Next»