more materials questions
Does anyone one know why this setup isn't restricting the emissive colour to the areas inside the UV map?
10 Replies
Because that's not how UVs work. And the saturate isn't doing anything.
While some workflows pretend it works differently they're special cases that in games actually require unique materials. How it usually works is a UV is the coordinate to look at the texture below, and the texture can either repeat or clamp the information when you go outside 0-1. So nothing you're doing in the material is going to change which polygons look at what pictures. You need to differentiate the polygons you want to be emissive from those you don't. This can be done any number of ways from an if, to math, to reading another texture. All with pros and cons. The simplest most performant material fix you could do is multiply the output of the white texture sample (or just replace that with a white color parameter) and for the other side of the multiplication do the sign of the texture co-ordinate Y value. Then feed the multiply into the saturate before emission. You may also want to multiply the output of the saturate by another material scalar parameter for brightness. A lot of emissions are usually stronger than 0-1
alright, so it's kinda moving in the right direction
I tried sign and sine, I assume I need to use sine because sign doesn't do anything
I tried changing the value on the "sine" currently at 1... and I get mixed results
thats the best result i've got so far
Would it be easier to just run the mesh with 1 UV map, add a texture, and bake an emissive texture?
Yes, it's easier, but the problem is the vram of a computer will get used up more and more quickly as you add more and more parts. So it's better to avoid it until you don't have much choice. That's why I said math is a good solution.
I meant sign. Sign takes the value of the value and if it's greater than 0 is positive, if it's less than 0 it's negative. So that should of seen those UV islands down there with Y co-ordinates of -2 or so. Then when the saturate node was applied the negative values should of been clamped to 0 overall giving you white for the upper UVs and black for the lower ones.
Sine takes the input of the value and maps it to a sine wave. Which is useful, but not for this. But by hammering the value with an accidentally useful value you ended up with something that worked. I would not recommend it as a long term strategy compared to learning the math at work 😄
actually the sine wasn't quite perfect. It's catching on the bevels