Advice on texturing

#1

Hello there,

so I’m currently modeling a house in Max and I have been stuck for some time now as I’m having a hard time understanding the process of texturing.

First off I would like to ask for an explanation. I have seen people use several different ways of applying textures on a model:

either they bring the texture to Max and apply it on a desired surface and adjust the scaling/tiling (video)
or they unwrap the model and bring the whole thing into Photoshop where they basically paint over the guidelines (video)
or take all the textures and put them together in a single image, bring that to Max and place individual parts on top of it (video)

What are the different benefits of each method and how does texture resolution fit in all of that? I assume tiling allows for a detailed but repetitive texture while the second method lets me put in more unique details, but restricts the overall resolution as all the parts are meshed in a single picture? The third one looks rather clumsy.

For my model I expect to use tiles because I was given a resolution restriction of 1024x1024 pixels.
Now if you take a look at the house that I’m modeling the large wall you see currently looks this way.

What do I do if I want to bring in all the details like the ivy and cables? Using one picture for the whole wall sounds like a bad idea, so perhaps dividing it up and using several textures would do the job? What would be the best way to tackle this problem?

I would be very grateful if someone could help me wrap my head around this.

Regards,

Daniel.

#2

You are restricted to 1024x1024 for a single house? What is the target engine/render? Modern characters have multiple maps @ 2048 or more and they don’t even occupy the screen space of a full house.

Resolution should be dictated by screen space occupied by object or rendered size if it’s not real time project

As for texturing approach:*You’ll need to be creative in how you build your mesh, the less geometry you have, the more you will need to use larger portion of your texture budget. If you cut the model in strategic places, you will be able to play around with tiling and re-use details (such as windows). The drawback of course is you will add more polygons but that house looks super low poly already.

#3

I’m restricted to 1024x1024 texture resolution and 2000 tris for the model.
The house will potentially serve as a part of a world in a driving simulator, so I suppose the detail doesn’t have to be that deep.

The whole thing in its current state:

Alright, so 1024x1024 should be for the whole house? At first I was confused whether it stands for that or a maximum resolution of one tile.
But how do you then control/manage the resolution when you are applying the textures individually as in the first video? Or do you just place them on the model and only then export them together in one file which will have the target resolution?

#4

I have no idea, you’re the one that has been assigned specs for the asset :stuck_out_tongue: if you’ve been told you are restricted to 1 map at 1024*1024 then you will need to slice that model up and be creative in how you tile the UV’s around, that’s the approach in the 3rd video, go with that.

Plan ahead as if you were building with lego blocks or modular assets if you prefer, which is what he did for that building, you can see the geometry repeated all over to get one giant building. The problem with that approach is of course the lack of unique detailing, the vines you mentioned for example, texture resolution should be a priority over unique detailing anyway.

To answer your question about the 1st video: easiest approach that would work (if supported by game engine) you would detach every polygon that shares the same texture map of the roof for example, once exported that chunk of the house would have the roof texture assigned to it and nothing else. Repeat process for every other map. So your house asset would become a bunch of sub-objects each with their own texture map assigned.

Another solution to overcome the problem would be counter productive and give crappy resolution:
You would have to remap those uv’s into something coherent in order to bake them down to a new texture sheet. You don’t want to waste time on that, it works for characters but not for modular assets and “creative” uv mapping.

Which brings us back to the 3rd video: it’s always better to fit your UV’s into an existing texture atlas imo.

#5

Right, sorry, that came out a little awkward, haha.
What I meant was the usual understanding of the word “texture” in that context.

Anyways, thank you very much for all your answers, Psyk0!
I imagine I will report again once I dive into it, but I already have a better understanding of the process.