I know I can in this example apply insert face then extrude, but in complex models this can increase the number of polygons, besides letting them apart makes the job easier.
this is totally fine. infact, a lot of the time this is preferred. however, uvmap space gets wasted on the overlapping bits which is important for baking.
in the end, geometry isnt really a big concern nowadays, unless your working on low end or mobile.
i have never run into a game issue with even some of the most wild shapes. its the triangulation that usually causes problems.
Depending what game engine it is you may get shadow issues if you go with separate meshes. Meaning that the shadow of the smaller bit may not connect correctly to its bottom of the shape which can make the shadow look detached. It is easy to try it in the game and if it works well then use it. Most of the time I’d not expect any issues.
Also I think if you separate the meshes in uvmapping then the vertices are split anyways so you don’t get any benefit with fewer vertices by having the perfectly joined together.
For static meshes, or a dynamic mesh that is mechanical, its totally fine. I’d advise caution when doing it for organic characters, as it can make rigging pretty annoying.
another aspect is z-fighting, 2 meshes with overlapping faces in the same plane, intersecting faces is usually o.k, as above but overlapping can cause issues, ensuring you have file structures and workflow setup for quick alteration of problem meshes on the fly is very helpful.
As mentioned above, view in game engine, if it works, it works, if not fix it.
upon rendering, everything is converted to triangles and fed to the gpu. this means the computer gets to pick which way the faces are cut up. manually triangulating gives you final say of the shape and shading. alot of the time this isnt really necessary for high poly, but i would say a requirement for low poly.
i think you would be surprised at what “professionals” like to try and get away with. i say test your game with whatever modeling methods you like, and see if theres any adverse effects.
Just to make it clear that I did not want to scorn anyone on the forum, on the contrary I know who there are great artists here and newcomers, like me. And when I told professionals, I referred to major producers.
The idea I had about the “big producers” was that they made everything perfect.
I will follow my model, and I will post here in the future and ask for your opinion.
Theres thousands of technic and other purely on the code of the engine API… from culling triangle ( for dont make gpus calculate invisible triangle ( not in front of the camera, back faces etc ), out of order rasterization (or rasterizer binning ) etc …
This is software developers, graphics engine developers in studios who take care of thoses optimization, not much the artists. ( who just create by definition, they can just be recommended some good practice if the engine require it ) … GPUs are working so much differently than 10years ago, let alone the APIs. …
Some example
Anyway, for a question like thatt, i will most goes to unreal and unity forums ask it…
