UV texturing/mapping noob help

[Jacfat]

Hey guys. Ive been working on understanding the pipeline for creating high detailed models for use in next gen games and I have a few questions that I havnt been able to find out through tutorials. The software I use is Maya, photoshop and zbrush so if you could answer in reference to them.

- Firstly when I UV map something this is done on the low poly model then I bring it into zbrush to generate a normal map with details. So when I return to maya I apply this normal map but my uv's remain low poly and dont reference the new details of the normal map or hi-rez. How to I get greater detail for reference when painting ontop of my uv's?

- If I am using zbrush and I end up drastically changing my models lowpoly shape and want to put the back in maya will the low from zbrush retain poly information? Is it smart to take in the new low poly from zbrush rather than sticking with your original maya low poly?

- Do i have to layout uvs befor I go into zbrush or can it be done at a different part in the pipeline.

- I am also wondering what software is used for texturing these nexgen projects is it mostly done in photoshop just painting ontop of your uv template.

-Also I am kinda struggling with the concept of making specular maps. Could someone give an example of making one? Is this also done in photoshop.

- Are the main maps used in next gen diffuse, specular,normal and alpha if required? or are there some other popular ones that are used to create next gen characters that I should be looking into.

Thanks very much for your help these forums are always great to read through for inspiration and help.

[Talon]

For baking, you don't need to have UVs on the high poly model at all - you only need to unwrap the low poly mesh. Your UVs don't become "high poly" when you apply a normal map (nothing does - that's the point of normal maps :P).

You can unwrap your low poly model at any point in the pipeline - it doesn't matter.

[Reverenddevil]

To answer a few of your questions...

You do not want to change the silhouette of your low poly model that you bring into Zbrush.. Get the low poly silhouette you want and bring that in and stick with the silhouette. So, get a really good silhouette then work on the surface detail in Zbrush.

The software almost every game artist uses to texture is Photoshop or an equivalent program. Yes it is all just painting on a 2D surface on your UVS. Some people use Deep UV paint or some other type of program that allows you to paint directly onto the model. Zbrush I think has a certain type of feature but not as in depth...


As for as specular maps go, just think of it as the more white something is the more it reflects light. An easy way to get something is to take your diffuse and desaturate it and then adjust the birghtness and contrast. Then to take it even further and get the best most professional results you should go in and paint out in black the things that don't reflect light or reflect light, such as carpet, some types of stains, some rubbers, etc.. Anything that doesn't get a reflection or shine to it when light is put on it. Then in white or variations of grays would represent how much reflection a surface gets when the light hits it.

As far as next gen shaders the main ones are

Diffuse= the color or normal texture of how something will look

Specular= Will determine the reflectivity of the surface properties of objects you are texturing...

Normal= Reading the color spectrum in the map to fake geometry. Giving a low poly details from a high poly, making the user believe something is more complex than it really is..

Emissive, Self Illumination, Glow maps= Pretty much names all used interchangeably. These are used to give objects or parts of the texture that cool glow look, like the glowies in Gears of War and all sci fi genres... Everyone loves glowies....

Alpha= Used when needed with any generation of games. Used usually for fences, shrubbery and intricate details that building geometry would result in to many polys or other types of effects...

Other maps for characters there is nothing different the same concepts apply as above... Unless There is something specific out there that I have never heard of..

Well I hope this gives you a little bit of a start and helps you to get a little better picture.. If you have anymore questions please feel free to ask....

[Jacfat]

Cheer Talon and Reverenddevil both great and very helpful responses. I have found the deeper I go into working out the next gen pipeline its get more complex but more exciting at the same time.

[thanoz]

Hey don't forget Occlusion map (even though it technically does end up being applied to the diffuse). Occlusion maps are used out the yin yang on all those WoW models. Might not be next gen, but it's certainly current gen.

Also another tip with Specular maps - on any parts that are skin, use a dark gray, run a noise filter over it, then darken and increase contrast until it looks like sparkly sandpaper. Try a render with that on and it will break up all the highlights on the skin really nicely like pores.

[Jacfat]

What do Occlusion map's do are they to do with color as they can be combinded with the diffuse map?

Also guys I have spent some more time poking around maya and zbrush. With UV's dont you require them to start making a normal map in zbrush as the uv's are what the normal map referers to in order to be placed over the model correctly like when u use zmapper it goes over the region filling out the details.

Also can someone explain why something would need multiple uv groups and how this would be organised. And in maya when you use that automated layout tool is that so the textures can be contained within a square dimension so texture sizes can be like 1024 by 1024 and 2096 by 2096 etc?

Thanks very much for your help guys hopefully i will have something cool to show in a few days. Currently I am still struggling in understanding many of the concepts a professional artist deals with.

[Reverenddevil]

As far as UVS for Zbrush, you will always need UVS so it knows where to place the information...

Occlusion maps are a good way to lay down lights, darks and shadows in your diffuse map. These maps are only used in photoshop as overlays or multiply over your diffuse texture. They help things alot.. You can also go in and hand paint the stuff but occlusion is a cheap fast way..

Usually multiple UV groups are used with characters, vehicles and terrain. Say for a character you can do the body at 1025 and the head 2048 etc... So, it comes down to making things more detailed with maximum texture space. So the head and body would be on 2 different texture sheets. Also for terrain you use multiple uv sets for detailed terrain..
So, i hope this helps explain a little more... Good luck...

[Jacfat]

ah yeah I see uv's are the foundation for laying all these maps. I have come to the realisation that normal mapping is next gen and only works well with a hi-polygon model for example here http://www.game-artist.net/forums/wo...rewarrior.html I can see his head is around 3500 polygons which is huge compared to the head i am working on with a meer 500 polys. When i tried using zmapper to create my normal map I was faced with this ugly looking beast.

Normal mapping only i think has really sucessful results with higher poly meshs. Is this correct? (because there are more points to reference the depth information from). Cheers for your continual help guys i think as I run through this process I will continue to ask questions when I run into new walls.

I also have a bit of a software question for normal map generation. Is it fine to make a normal map in zbrush with zmapper? or is it prefered to use "transfer maps" in maya to create a map? will they both result in similar renders?