UV, image size and polygons count for correct sculpting and baking

thinkinmonkey

Hello, I have a doubt about uv and the right polygons count for sculpting/baking correctly.

I want to share with you what I think and see if it's correct or not.

if we have an image of 1024x1024 pixel, theoretically it can store 1,048,576 polygons normal map value, if we unwrapped every single polygon to every single pixel.

That means that, if I import a game asset in zbrush, it should be better to subdivide that mesh without exceed that number.

Basing on that, if I have a character or a baseball bat, for example, I unwrap the asset and it covers, for example, the 50% of the uv space, doing a quick evaluation, I would have 524,288, that means that, roughly, I shouldn't exceed that number when subdividing in order to have the roughly correct baking for that image size.

I mean, at least, I could do a more subdivision just to smooth.

Does it make sense or it's just unuseful math?

Thank you

icegodofhungary

Obviously the goal is to unwrap your models in such a way as to make use of the most UV space available. And I think a general texel density calculation could be helpful in determining what size of texture to use. But I haven't seen anyone draw a link to the amount of polies in their high-res mesh vs the texture size on their low/mid poly game-ready mesh. You add as many polies as you need to get the detail you want, regardless of what size the final texture is. Even if you're doing a 512x512 texture for a trash can, your high poly mesh might still have 1,000,000 polygons. You wouldn't be saving any time or anything trying to find a correct ratio of the high poly count to the texture resolution.

thinkinmonkey

Thank you, icegod, I thought it was something to be used as guideline, but I think experience knows better.