Rules of thumb or limitations of Blendshapes vs Joints and SDK in Unity?



I am looking for a bit of input and suggestions on best practices or things to think about!

I am getting ready to rig a single character that will have its full body facing camera. Possibly some mid shots.

He is an expressive cartoony character. I would like to nail his mouth corners and his eyebrows especially.

I am familiar and comfortable with setting up hybrid blend shapes/joint face rigs for characters. I am more comfortable with blend shapes than joints for things like the mouth corners in particular.

My character’s body is a single mesh.

I rigged this character and rendered him in vray previously. I duplicated and separated the head to create a bridge and a series of blend targets off of that.

I have been getting errors when I try to FBX export my separated head method of blendshape targets that my vert counts don’t match. I am however, confident the vert order matches. If I want to create blend shapes do I have to create full body blend shapes for each target? Will this slow down my unity scene?

So I am also looking at the option of setting up a SDK joint based face set up vs. blend shapes.

I am also vaguely familiar with the idea that I need to balance the amount of joints I use, and possibly the amount of SDK curves I create…but I also imagine that the amount of blend shapes especially if they need to be full body blend shapes will add up. (no pun intended)

Are there any contemporary resources or any advice/rules of thumb for choosing between blend shapes and SDK driven joints? Then determining budget of joints I should be working with? Or some rules of thumb or best practices of utilizing blendshapes into unity?

Thanks in advance!