So Im going to start the reading and video watching asap for the project im working on with per zbrush. I still want to clarify a few things. Does the object you import into zbrush need to be one mesh? Meaning no separate parts? In example shoe laces on a shoe.
Now, if Im planning to make a additional low poly version of the character at the end. Do I need to bother adjusting uvws before importing into zbrush? I mean since I can then when I export back into max just burn both the normal map and color map to the low polys layout. Or should I instead import the low poly into zbrush and burn the normal and such inside of it?
I ask in part because I dont have that polygon reduction program you all use to keep max from crashing with the huge polycount zbrush makes.
Im still confused on displacement maps. Why would I need one for a real time engine? Are the newer engines subdividing your base model real time to contour to a displacement map?
Also, Im making this character so that its facial features will animate. Do I model the face then with the eyes closed initially and the mouth slightly agap? Do I bone a tongue also for a real time engine? Anyone have a sample charcter that has a more advanced facial rig [meaning it can show full emotions] that was made for real time that I can open and study in max how it was constructed?
Im making this looking at UED3 if its anyhelp. Which then leads into anyone have any idea of how to split the textures? In UED2 you split the head from the body, but I have no clue what Epic has in store for us with how the characters are built. I plan on downloading rorch's thread on the subject, but he was only covering environmental assets.
EDIT: oh wait, I got displacement map confused with a height map.
Replies
2. I wouldn't worry about UV's when zbrushing if you're going to make a new lowpoly mesh at the end. Regardless of whether you bake normals in Max or Zbrush, you can use a new object with new UV's to bake normals onto. You'd only need to UV-map the zbrushable mesh if you wanted to texture the highpoly, which for games probably isn't necessary.
3. You don't need a displacement map for an in-game object. Normal maps alone are fine. Height map isn't really needed either. You can use a fine-detail greyscale map (these are used in doom3) to overlay onto the normals to add detail to skin or cloth. Not absolutely necessary though.
4. Face rig depends on what game it's going into ... i don't have experience in this area, but i'd say anything you want to move should be boned. Add a bone for the tongue if it's gonna need to move for lip sync stuff.
Modelling with the eyes up or down shouldn't really matter, you could do it either way and just work it out in the rig. Mouth slightly open is probably a good idea just so you don't run into trouble when zbrushing.
I'd imagine you can split up textures pretty much any way you want since you'll be defining them yourself anyway. It's probably safe to do a map for the face and a different map for the body, but i'm sure it wouldn't be a problem to put them all on one map, or to use 3-4 maps if necessary. 2 is probably optimal though.
Per 3, why is there so much discussion on this subject then (displacement)? Also you meant a standard bump map for the details? Did doom 3 have a separate texture for the clothes so it could repeat that texture across without hitting skin and such?
I ask on the textures because in order for team colors and the big head mod to work you had to have the textures split in that manner for ut2k3/4.
Thanks for the info Mop.
Also you can use lower-level subd meshes from zbrush as a base for a final in-game mesh.
Since you know Pior had a tutorial on it, why didn't you check out his site? The tutorial is here - http://www.pioroberson.com/images/minitut_topology_worflow.jpg
I'd imagine most discussion on the topic of displacement is actually referring to pre-rendering stuff out of zbrush (like making a displacement map to apply in max and render a scene)... that sort of stuff isn't used in games AFAIK.
And yes, a standard bump map is used in doom3 to add fine detail to the maps ... i'm not exactly sure how since i haven't researched it much but i imagine you could find out easily enough on iddevnet.com or doom3world forums.
As for texture splitting, you answered your own question there
If I plan on baking in lighting, then should I get the displacement map also?
Now for the facial animation part.
[ QUOTE ]
Regardless of whether you bake normals in Max or Zbrush, you can use a new object with new UV's to bake normals onto.
[/ QUOTE ]
holy shit, can I do this in ZB too? I have always thought that you have to a basically finsihed lowpoly model in quads and with uvs to use ZMapper and then sculpt everything in one model.
So I can do a midpoly in max, export to zb for highpoly lovin' then export the finished max lowpoly and create normalmaps with the two meshes (low and highpoly) similar to Max render-to-texture feuture?
Mop, could you please teach me how, if it's actually possible and I'm not just reading too much into this .
If you don't have the help file you can get it from zbrush forums, just do a search over on zbrushcentral or something.
Before that I just left-click somewhere on the canvas and move the mouse to rotate around my object. Once the problem comes us though, I left click and it either makes a new Zsphere or does something different.
I've been trying to find some basic Tutorials on Zbrushcentral.com, but I can't find something that really covers the interface.
Opdate: Never mind, I had a friend at school who's almost graduated help me out. I'm still just learning, but the interface because a lot clearer than trying to figure it out myself. Now all I need is practice, but I couldn't believe the cool stuff it can do.