Environment Workflow

ayoub44

hi guys

i want to ask you about workflow on creating environment
so i see Nate Stephens dvd about creating environment great DVD ... and he is creating a next-gen environment but he just go with modeling . my question is why he is not creating hipoly mesh ? he is not going to use normal map ? it's next-gen environment .
so my point is evry mesh in the scene do we have to make for it hipoly to get details ? every mesh should have a normal map ? and when we use PS to create normal map and not baking it ?

soory for the long question

m4dcow

I've watched his DVD and he talks about using tiling textures on alot of spots. He does have a little section at the end of the dvd about creating normal maps, so while it isn't really explicitly stated that normals maps are expected to be used I think it is obvious.

Anuxinamoon

Making enviros is different to the traditional character work flow. Being of a character background and moving to enviros for more of a challenge, making highpoly assets for all of your environment pieces can be a waste of time especially if you are utilizing tiling textures and trim strips.
Sure there are unique assets that will need highpoly source meshes to bake from but also remember some environment pieces (trees and architecture being some examples that spring to mind) don't need to get UV'ed until after the textures are completed, using existing texture libraries to make extra content faster.

Creating these texture libraries will sometimes involve making a high poly asset to make normal and diffuse data from. But other times you can get away with making a simple height map and using a program like Crazy Bump to create a normal map from. This is especially useful if the texture has been partially or fully done and you want to save a bit of time (like if you were photosourcing textures.)

You want to be able to make good quality assets smarter. Increasing productivity without loosing overall quality in the same amount of time. Making a highpoly for everything looks cool but if you can get the same or better result for cheaper; productivity wise and tech budget wise then you can understand why highpoly models are not needed for everything.

Does this make sense? hahaha I'm so bad at explaining things xD

ayoub44

you right but how can i get the same result without making hipoly and loosing quality , if we have a simple object . we can just baking normal map just for having some cool and smooth edge . ? thanks you

m4dcow

Well Anuxinamoon is saying that in production you don't always have time to make a high poly version for everything. There are certain things that crazybump would be mush faster at doing and probably provide a better result (low frequency detail ie:concrete, sand etc...

Even if you do have time to do high poly versions of certain things, it would likely be things like trim stips that would be re-used around the environment and even then some of those tiling types of things could be doing very quickly in photoshop or crazybump.


Check out this thread:
http://www.polycount.com/forum/showthread.php?t=81808
I don't know how the artist specifically approached the scene, but it uses alot of tiling textures for most of the environment, and by the look of some of them they may not have been baked, but the environment still looks kickass.

ayoub44

thanks man now i understand ... thanks you so much for the link too

passerby

also to save time think about how you can re-use details or even parts of models.

say you do high and a low poly for a stair case, and bake it down, and the front of the stairs has a lot of nice detail baked from the high poly. you could always uv map a plane to that same area and use it as a trim piece somewhere else in the environment.

that allows you to quickly make a new piece for your modular set, that uses no extra UV space, and it can help tie things in with your scene. since it is also possible to be modular in how you make your modular pieces.

also with props you already got done and baked down, play around with there scale, and maybe adding bend deformers to them to re purpose them.

if you look at some of Kevin's work in UT3 and GoW you can see after he bakes down from the high he will add bends and rip apart the LP and add parts to other props to get more use out of it. there is one pillar in UT3 that he uses as a pillar, than he ripped the side off it for a trim piece and he also bent it into a half circle and used it as trim for a door.

Bruno Afonseca

this should clear up a bit:http://thiagoklafke.com/modularenvironments.html