4 bit texture bug

[Ian]

Maybe I am missing something obvious, I don't know

I changed 8-11 to

Code: Select all

         for (xi = x; xi < (x + width); xi++)
         {
            texel = src[yi * 2048 + xi] & 0xFF;
            c = (texel & 0xF) * 17;
            scratch[i++] = c;
            scratch[i++] = c;
            scratch[i++] = c;
            scratch[i++] = c;
         }

Fragment shader to

Code: Select all

   "if(textureInverted) {\n"
      "tex1Data = vec4(1.0) - vec4(tex1Data);\n"
   "}\n"

[HarryTuttle]

Maybe I am missing something obvious, I don't know

I changed 8-11 to

Code: Select all

         for (xi = x; xi < (x + width); xi++)
         {
            texel = src[yi * 2048 + xi] & 0xFF;
            c = (texel & 0xF) * 17;
            scratch[i++] = c;
            scratch[i++] = c;
            scratch[i++] = c;
            scratch[i++] = c;
         }

Fragment shader to

Code: Select all

   "if(textureInverted) {\n"
      "tex1Data = vec4(1.0) - vec4(tex1Data);\n"
   "}\n"

The fence in the precedent post is type "6", don't know why it was inverted... anyway, the alpha channel for 8-11 is always inverted, but rgb values only when you, rightfully, set the inverted bit by means of "translator map offset = 2"

[HarryTuttle]

In pseudo-code:

Code: Select all

if (textureType > 7 && textureType < 12) {
  tex1Data.a = 1.0 - tex1Data.a;
}

if(textureInverted) {
  tex1Data.rgb = vec3(1.0) - tex1Data.rgb;
}


[Ian]

I'll try that

[Ian]

Well harley looks correct, but still no black around the text

[HarryTuttle]

strange... I've just followed the current, official build solution regarding types 8-11, with only one variation: I changed alpha channel interpretation from "0" or "1" (like a contour texture should be anyway) to the full 16 grey-level, but in both solutions alpha is inverted.

[Ian]

Dunno, seems to look correct now. I forced a rebuild ..

[HarryTuttle]

[model123]

Great progress!!





Thank you for a great job!
Supermodel developer