A Quick Look at My Personal Model

That rabbit-boy model from before is basically finished now.

Since I don't really need a lot of body-language stuff for streaming, and the outfit design doesn't really call for showing off the body anyway,

this round I focused mostly on the facial details.

I also finally dusted off VBridger, which I basically never use.

In terms of performance, the tracking accuracy went up noticeably, but there's still a good chunk of mouth parameters that need tweaking.

Also, I'm just naturally pretty deadpan-faced myself, so this probably won't help much for my usual chill streams — it still takes a decent amount of active expression control. And honestly this approach doesn't really suit gentler characters either — it can end up feeling a bit OOC, so it's something you'd want to use carefully. Other than that though, the results have been pretty great overall.