Possible improvement with cloth bones (Patreon)

Hi everyone,

If you've ever wondered why I use cloth bones instead of rigging stuff directly to the body bones, the images above should give you a good hint.

What's following is a bit technical so if you aren't interested, you can skip to the last paragraph.

Images are of course an exaggerated example to make a point but basically : it kinda makes sense that the hanging cloth would move with the legs in some situations. Problem is, when you attach/rig the cloth to the legs, you can't choose which rotation you want or don't want, you take it all which can result in what's shown above and I hate it with all my guts. I'm sure you could fix it with a lot of time and sculpting but that's still a terrible starting point.

My current solution is to have cloth bones specific for parts of the outfits that shouldn't follow every body move. It solves one issue but results in another problem : when you move the leg, the cloth bones don't follow and you need to adjust things manually...still better than first solution but not ideal either.

I might have found a way to help with this last issue by adding bone constraints to cloth bones. This way I can tell the cloth bone to copy the leg rotation and I can also limit rotation to certain directions. So for the front cloth, the bone would only follow if the leg moves forward, not if it moves sideways or backward. This way, we keep the cloth bones for fine tuning if needed, but still have some movement when it should.

I've tested this a bit today and it seems to work alright, although there's some issues.

The main one is that it's awfully tedious because I have to add and tweak several bone constraints for each bone manually...and bones' axes aren't necessarily oriented as the world axes so it makes my head boil to convert rotations between various orientations. 

In conclusion, bone constraints are a possible way to improve the current way of doing things but I'm not sure if it's viable in the long run given the complexity and time required to set it up. I probably wouldn't dare deploy it on Kitana for instance. I wanted to share this because someone might have another idea.