First Picture: where a cube is unwrapped as you can see there are squares in Color Grid (512x512). The level of detail that a Cube would get is pretty high.
Second Picture: Where a cube is marked with the seam to get unwrapped into 6 square using the same Color Grid (512x512). The level of detail would be much low compared to the first picture.
Now, If the cube with 6 Squares in UV map is to have the same level of detail which is 512x512, then it would need to have 2048x2048 (4 times the original size, 512x512).
that should give you the basic idea of where this is going. Now What if I have a complex shape like vehicles or any objects where I unwrapped them in a UV map with multiples islands. The question is how do I make them have the same level of detail as the first picture, cube with 512x512? Again, like a cube with 6 islands or square will need to have 2048x to get the same detail.
Is it possible to get it done? I would like to have my work to get the same detail with 512x512. more or less or exactly?
Edit 1:
This is related to Texel density or pixel density.
Edit 2:
Okay, I’ve done my research to learn more about texel density. There is an add-on called “Texel Density Checker” which allow me to get what I am looking for. I would like to learn more about UV mapping and the size of texture Image I should be using.
For example, I have made multiple islands into a size of 128x128 and put them in a texture image consist of 2K x 2K. it has some space leftover where I could put the other islands from different objects into the same UV map/texture image where the Multiple Islands were placed in. These islands can be also a 128x128 (the size can be varying)
Is this fine?
I am aware that the higher texture image size you are using, the more memory usage that computer will have to use to render the object with a texture image. You can see I’ve done a lot of research to learn more about. I am opened mind… If there is any good advice that I should know, then please share…
The simple economy of UV texel dentity, is memory use vs visual fidelity, which means squeezing the most resolution per your uv islands.
Texel density example: in game buildings, you see a lot of STACKED UV ISLANDS for repeated parts, as in positioning similar uv islands on top of each other. like windows and roof tiles. This is to maximize texel density, to get the max visual fidelity for object/texture map. 1K, 2K, your choice.
The second thing you’re seeking is Texture Atlas. Yes you can have one texture map space for multiple objects - in most 3D creation tools and the majority of game assets. Super memory saver.
I’m a ZBrush user still new to Blender’s UV process, someone else can advice you on texture-atlasing or UV space sharing between objects or object parts in Blender.
Don’t worry about it. you are just getting the discussion started. So Texel Density example is good one. that got me thinking how it should be used.
I am very familiar with Texture Atlas… You could say I am an expert in modeling and texturing using node editor in Blender. BUT i do not have enough knowledge to know more about texturing and how it should be setup for gaming.
Option A: What if there are 3 objects that have: the body and 2 wheels of a vehicles. If you unwrapped them that they have their own UV map separated from each other. That would take up lot of memory usage because there are 3 texture image, right?
Option B: Same with 3 Objects, Texture Atlas can be used to unwrapping all of 3 objects in the same UV map’s space. the islands of wheels are not in the same position because they have texture where they are dirt, rusty or old. So 2 wheels are separated. The texture size can be a little bigger than Option A. This can save a lot of memory usage because there is only one texture image? Right?
not much more I can add this is from the blender wiki on UV mapping
may be you can find it and read other things about UV!
I think it calls this Anti-aliasing
