When is it necessary to have a continuous mesh over separate pieces?

Question

Everywhere I look people say that animated models need to be one continuous mesh with quads. A lot of tutorials of people who are doing hard surface modeling via poly modeling vs sculpting often model there final object with several different pieces. Like if it was a helmet with rivets they'll model the rivets separate and just leave them on top. If you want to animated the helmet on a figures head do you have to join all of it into a solid mesh? if you do that do you re-topology it the same approach as a sculpt?

I ask in reference to a medieval great helm I'm modeling. it would be easier to model the rivets, and the cross (separate pieces that is riveted around the eyes and seams. I want to be able to 3d print it and but it on a character to be animated. Does the entire helmet need to be one solid mesh? I don't want to sculpt it.

Answer 1

To be clear: First you say

Everywhere I look people say that animated models need to be one continuous mesh with quads.

Then you say

A lot of tutorials of people who are doing hard surface modeling via poly modeling vs sculpting often model there final object with several different pieces.

Simply understand that it's not necessarily a contradiction for people to be saying these two different things, because not all modeling projects have the same goal. I suspect that many of those you've been hearing are simply making assumptions about what other people are going to be doing with their models, and so maybe they are not entirely clear about their own goals when explaining their methods, so that others will understand what those methods are best suited to accomplish. Sometimes we take too much for granted.

Hard surface modeling can use various separated parts, just as you say--and even boolean them together without worry--because the end result will not need deformations. When you hear people say that models need to be one continuous mesh with quads, they are likely assuming the goal of organic modeling, such as for characters.

Once this is understood, the course becomes rather clear, and there's no need for concern. For a helmet, you could probably do whatever you wanted. The only real issue I can foresee might arise from the fact that you want this model for 3D printing. That is going to require you to take more care than might otherwise be necessary, but I can't say in exactly what way, as I am not an expert on modeling for printing purposes. It might be a good idea to research this topic specifically, since the cost of physical material is at stake.

Answer 2

This was a big issue for me too, since the beginning. Still I'm not sure but what I understood is that if you want to just render something you dont care to join,if you want to print a model as it is you want to join and have perfectly smooth face without using shade smooth from object mode.

If you want to make a mould of a model to be printed I think its ok not to be joined.

For games also need to be joined as much as possible,there you export/use fbx.for animation you need to join the body for example and then parent other objs like helmet, eyes, particles etc., but parts like body, needs attention. The mesh should be detailed, with many verts, because when you animate something like a movement, you use weight paint and uvs etc,if don't pay attention it will be just a disaster.

Answer 3

Games - joined for organic shapes. Non-joined for stuff like robot's arms. Try to join everything that is supposed to be one object and separate pieces that logically would be separated.

3D print - everything perfectly joined no intersections

Image/Animation, do what you want tbh. Intersect, don't intersect, your chair is clipping trough a wall? No problem!