Update on my blender importer fork

Discussions about MHX2 and other plugins that are developed outside the scope of MakeHuman. Note that bug reports should go to the respective code projects and might go unseen here.

Moderator: joepal

Update on my blender importer fork

Postby LynnSpyre » Fri Feb 18, 2022 8:13 am

Hey all,

So I decided to take it upon myself to do some vtuber specific work on the makehuman mh2 importer for the very specific use case of vtubing. What I want to create is a generic importer that'll the auto-rig faces, so that you can go straight from your blender import to Unity, if you wanted to. I'm planning on implementing the entire Arkit-2 expression catalog, the base emojies, the visemes, a handful of targets for tongue tracking, and aliases for vrchat. At this point, it's more figity leg work than anything too terribly difficult to do. And the larger community of vtubers and metaverse app developers need more open source options. Makehuman is perfect for this.

First thing I did was figure out the nuances of the targeting system. How it works, and how to convert targets that come out of makehuman into targets python can use. And I wrote a quick parser to bridge the gap in the formats between the target files and the faceshapes.mxa file. Had some issues not knowing the difference between a base target and a diff target, but I figured it out. Got some revised visemes in, which follow the chart a little more closely than the ones that are there now (a couple of them are broken, but I'm working on it).

And it was going pretty well, actually.

I was able to get eyelashes and eyebrows to rig. But I'm having issues with the eyes and tongues. At least programmatically. I'm having trouble getting targets with these movements to translate down to Blender. The keys come through the import, but the poor model doesn't respond as expected. I've tried messing around with different topoligies for the eyes and tongue.

My thinking now, is that I need to do the shape key logic on those parts, as the system is selecting them. It's all text, so I should be able to diff them out of the bigger target files if I need to. I think I'm just somehow missing the mesh selection. And consolidating body parts does help in other areas, but here it just makes things unresponsive.

Anybody have experience with this?

Thanks
LynnSpyre
 
Posts: 8
Joined: Mon Feb 14, 2022 3:27 pm

Return to MHX2 and other plugins developed outside MakeHuman

Who is online

Users browsing this forum: No registered users and 1 guest

cron