UV curiosity

almahiedra

I have a big curiosity about why genesis and and genesis 2 female (I supose G2M too) have a unexpected behavior (at least to me) when I "color" them following uv values, specifically u-coordinate (I supose v-coordinate is the same). Victoria 3 and Victoria 4 behave as I expected.

In shader builder I made a shader which color in red the u coordinate if this is less than 0.5, green in other case. In the first image are render with the shader G2F in bow skirt dress, a hair, a sphere, a plain, V3 (left) and V4 (right), all but G2F behave as I expected, G2F is a mess. I want understand why.

In second image s is used instead of u. This it is the reason for which I wanted UV variables, because ST are a problem, though waited.

*shader with u-coordinate:

color red;
color green;
green = color "rgb" (0,1,0);
red = color "rgb" (1,0,0);

if( u <0.5) {</p>

Ci = red;
} else {

Ci = green;
}

Richard Haseltine

I can confirm that, but I don't know why it happens.

throttlekitty

Perhaps it's similar to the issue I ran into with the normals shader. Try turning off subdivision for G2F?

almahiedra

throttlekitty said:

Perhaps it's similar to the issue I ran into with the normals shader. Try turning off subdivision for G2F?

Yes. The first thing that I did. But u-coordinate have others weird behaviors. result for u > 0.5 is equal to result for u> -1. I suppose that I'm not understanding how the UV coordinates are numerically distributed on the surface.

throttlekitty

Odd. UV-grid coordinates start at the lower left, and increment U to the right, and V upwards. They aren't mapped in 3d space on the surface, if that's what you're asking- each vertex has their XYZ as well as UV values.

I don't know what your end goal is, but why not use a red/green texture split down the center?

almahiedra

throttlekitty said:

Odd. UV-grid coordinates start at the lower left, and increment U to the right, and V upwards. They aren't mapped in 3d space on the surface, if that's what you're asking- each vertex has their XYZ as well as UV values.

I don't know what your end goal is, but why not use a red/green texture split down the center?

Because the goal isn't split in two colors. Goal is understand ST-grid, UV-grid and points behavior. Yes, your answer is what I needed, now genesis/g2F UV behavior makes sense. Thanks.

almahiedra

Almost what I wanted. It is not perfect but at least it works better than UVs, it is a transform in object space. The problem here is object space doesn't follow the "route" of the body, but the "route" of the "box" that contains the boby.