One Whole Mesh or Broken Up Combined Into One Mesh

Inhert

I have been wordering what is the best way to create a final low poly mesh for riggers and animators to work on in terms of either retopoing one whole entire mesh or combining broken up parts into one mesh? I was doing a project once where I had separate meshes combined into one mesh and when it came time to rig/animate it they problems getting it to animate right.

Should I retopo an entire mesh without any separate (I think the correct term is "floating" meshes) meshes or can I have "floating" meshes combined into one big mesh for the riggers/animators to work on? This is something I am unsure of doing everytime I retopo a low poly mesh for rigging and animation.

rollin

depends on what it is and for what engine

If you have rigid objects which are connected to another rigid object then it's ok or maybe even necessary to connect them

If you have rigid objects which are somehow bound to another rigid objects and they can move in some way independently it might be good to have both models separated.

In a production you would ask the animation department to give examples of what they like and what not and check with the technical department if thats ok or not

antweiler

I would keep meshes that are naturally connected together (like a naked body) combined.
For modeling its mostly convenient to keep parts separated, so why not work in that way, and before heading it over for rigging combine everything and keep the separated model in the scene for tweaking and corrections. Advanced Engines like Cryengine or UDK can easily import separated meshes as one (combining them on import). For rigging its mostly easier to have separated meshes, especially for skinweighting.

Inhert

Thanks for the reply guys. I'm thinking I should just retopo the whole entire mesh as just one whole mesh. The character I am currently working on is a female character with a prison jumpsuit and gloves that have spikes on the knuckles. She also has shoes on with her arms exposed since it is a short sleeve jumpsuit.

I spoke to one of the riggers on the team for the female character I am currently working on and he said he would prefer it to be one entire mesh instead of one whole combined "floating" meshes. He said that if the model had "floating" meshes he would have a hard time finding hidden vertices for painting weights since the "floating" meshes are overlapping each other. The engine we are using is the Unity engine if that make things any clearer. I don't know much about how game engines work since I mostly model out stuff for the team and just hand it to someone who specializes in their own respective fields.

I did do a soldier model in the past for a game mod and I combined the "floating" meshes into one whole object but the person who was trying to paint weights on it and then tried to animate it had trouble making the animation to look right. I think it was because the arms and fingers were "floating" meshes that were suppose to be one continuous topology with the shirt and gloves but I didn't think it would matter at the time.