Hi everyone im having a problem with the resolution of when im creating a texture, primarily a bump map. I paint in my value in projection master (PICTURE 1) Pick it back up and i get the low resolution seen in (PICTURE 2) resulting in an ugly and nasty looking bump (PICTURE 3) im using a 4096 texture and used auv tile set to layout my uvs. i hope there is a simple explanation on how to fix this. Thanks everyone
What size is the model you’re painting on?
First, the dropped model is smoothed. The picked up model is not. This could account for what you’re seeing. Turn off Quick 3D Edit and set Tool>Display Properties>DSmooth to 1. Does the texture clean up? Projection Master assumes that in your final render you don’t want to see a bunch of polygons, but rather want a smoothed surface. So it projects texture based on that.
If that doesn’t clear up your problem, it’s time to get a bit more technical on you.
How many polygons did your model have when you did the UV mapping? The more polygons, the smaller the squares will be on the texture, and so the fewer pixels can fit on each one. This in turn can make it harder to get really crisp details.
Also, as far as I can tell you’re trying to paint VERY fine detail relative to the overall size of the mesh. Let’s put it this way:
Imagine a cube that is 100 polygons x 100 polygons in size. That is a total of 60,000 polygons, which comes to 244x244 on a texture where 100% of the available UV space is used. That means that each polygon could have a maximum of 16 pixels’ worth of texture from one side to the other. So anything that takes up less than 6% of the width of the polygon is not going to transfer cleanly to the texture. The detail is just too fine in relation to the surface area of the model.
Now that example is pretty simplified. Most UV mapping does not actually use 100% of the available UV space. (This includes AUVTiles and GUVTiles.) In addition, more polygons in your model would enhance the effect, and your mesh looks fairly dense.
my base mesh in maya is 17,527 polys/Faces in z brush it was divided 5 times bringing it to 4,486,912 polygons. i would like to divide it again but z brush is saying that i cant go over that amount, and the slider in my preferences is set to max which is 9 I believe. the first method didnt clean up the texture and i understand what you were saying about the uv spacing, but i would still like to figure out a way to make this work. thanks alot again for the help.
Ok, back to basic math here:
Your model is 4.5 million polygons. That means that AUVTiles will give a layout that comes to 2118 polygons across by 2118 polygons down. Not all the texture space will be used, however, since there will be some gaps between the polygons of the low resolution model in the texture mapping. But to simplify things for the sake of argument, let’s say that all the texture space IS used with no gaps at all.
4096 width divided by 2118 polygons means that each polygon gets less than 2 pixels of texture space.
The details that you’re trying to paint are only 3-4 polygons wide.
This means that on the texture, they translate to less than 7 pixels. In the actual UV mapping of the model, it’s probably closer to 3 pixels.
Do you see why it’s impossible for the line to transfer perfectly? The detail that you’re attempting to paint is simply too small in relation to the surface area of the model, even with a texture of 4096x4096.
Now here’s my question: Is your render REALLY going to be zoomed in that close to the model that the detail would even be visible? Hold your hand about 6 inches in front of your eyes. See all the fine lines and details? Now move it to arm’s length. How many of those details are still visible? If you were texturing a hand, and that hand was only going to be rendered from three or more feet away, would you bother with painting all the pores and other fine details? You’d be crazy if you did. Going from the examples you posted, if your model were a head then the detail you’re painting and the camera’s proximity to it is sort of the cinematic equivalent of zooming in on a zit.
Same goes here. If the camera will never be close enough to the surface of your model for the fine detail to be seen, then don’t paint it! Why knock your head against a mathematical impossiblity?
On the other hand, if the camera IS going to come close enough, you have another option. It’s called LOD (Level of Detail). Basically, you have one model which comprises your entire figure and is textured with a reasonable amount of detail. Where the camera will zoom in close enough that the level of detail is no longer sufficient, it’s unlikely that your entire object will be visible anymore. So now you switch to another model or texture which only covers the portion of the object which will be in the camera at that point. (In other words, either a partial model or a texture that is mapped to only a small portion of the model’s total surface.)
THIS texture is the one that you will paint with finer details. And because the texture is now much smaller in relation to the figure’s total surface area, your details are now going to be much larger in relation to the texture size. In short, you will be able to get the kind of detail that you’re after, should you actually need it.
But i like painting pimples and pores… dohh… haha i get what your saying and yes i tried painting in the large general color blocks and fromt he camera view that i will be setting up it should work just fine, thanks alot for your expertise and help!!! i really appreciate the dedication that you have to the Q and A forum your def a ZB guru if i ever saw one!