When making mechanical structures, I often put primitives inside one another to test out shapes and make up general forms. Then when I move on I make duplicates, detail the heck out of them, then retopo over the top of that.
If the detailed objects aren't MUCH different from the original shapes, like screw holes, screw heads, vent slits and other smaller details that won't be observed from a close distance, is it worth me doing retopo when the primitives are there in the first place?
What I mean to ask is better demonstrated with a picture, to follow shortly.
So, the first is just the two basic primitives put inside one another (Ooooh, err, kinky), the second is a triangled version of the two primitives, no other changes and the third is a full surface only retopo, no internal faces at all.
What I want to know, is which is the better for....well any kind of CG work? Which works best for game design? Which works best for CGI animation? What are the drawbacks/advantages?
EDIT: I just realised my retopo of the 3rd may not be the most simple way to do so, I am a novice at best, so please exsqueeze me if it's not the best and painting the wrong picture in this situation. It's a case of not knowing better.
I think when you do something that have 37 faces and ~20verts it doesn't really matter. But there's one thing that you need to consider - textures. One time separated things will be more comfortable to texture, sometimes you need to do it as a whole object.
And you need to remember of UV unwrapping - always minimize texture count.
Good question! Any object you import into a game engine will be triangulated anyway, so as far as games are concerned, #1 and #2 are identical. It's easier to edit as #1, so I'd recommend keeping it like that for now. Since it won't matter if your geometry is intersecting, and it's best to limit triangle count, I'd go with #1 over #3.
Yeah, I'm just wondering which out of Object 2 and 3 would be better though. I mean Object 2 may make texturing more difficult, but it would save on face count values but conversely, Object 3 would make texturing easier but cost more on face counts.
Which should I use objectively? Because on a larger model with a lot of details similar to this, saving maybe 40-50% faces would strike me as the obvious choice - unless there's another reason not to work that way. I mean I'm confident I could still make use of my UV space well with Object 2 by simply moving one of the partially obscured faces so that the obscured part would move off of the UV plane. Object 3 would be easier to make UV maps for sure, but that's just more time as opposed to better performance.
What I mean to ask is - which approach is the best, which is the most acceptable as an industry standard, and which would benefit a project the most, be it CGI art, animation, or Game models?
Thank you very much, I thought it would be something like that. This'll save a lot of time on modelling. Is there any reason whatsoever to use something similar to object 3, a case where a full retopo would be the norm?
Or is going from sculpture -> retopo the only reason to retopo like that?
Is there any reason whatsoever to use something similar to object 3, a case where a full retopo would be the norm?
Only two reasons to use #3 come to mind:
krajca mentioned #3 will save texture space, but for games the amount of polygons you save will make it worth it to use #1. Hope that clears things up!I think about another one reason to use #3. I think you can have some lighting problem with obstruct objects like in case #1.
For me as a unity developer this is just too abstract to judge and this "primitive" case can't be used to create general rule. In bigger models you can even have less vertices in case #3 than in #1 because inside part of models can be huge. And quantity of variables is quite big - amount of work, scene, number of appearances in the scene etc. Not to mention function of mesh itself: if it's foreground or background object, if it's animated and so on. There are some basic rules like all of rules mentioned before (less textures = better, less vertices = better etc.), but rest is project depend.
