openseeface-gd VS VRM-Addon-for-Blender

VSeeFace hasn't been made for Linux, but OpenSeeFace, the face tracker, is open-source and runs on linux using Python. And we use it to use other face tracking apps relying on OpenSeeFace, like Vtube Studio, Pupeteer / OpenSeeFace-GD

Contrary to what others are saying, you can load things from outside the res:// or user:// directories in an exported project. This project loads assets from a resources folder at runtime (and this is verifiable by the fact that assets can be modified after export).

This might be a option for me: https://github.com/you-win/openseeface-gd

It's not mine, exactly, but I got permission from the dev to package it into a flatpak. The dev themself stated that if it was easier to package it as a flatpak, they would do it themselves, so I am trying to make it so all they have to do is drag n drop the OSF-GD files in and build the flatpak. Link is https://github.com/you-win/openseeface-gd

OpenSeeFace-GD is an open source multi-platform Vtuber rendering program written in Godot. It has pre-compiled builds for Windows and Linux but is 100% compatible with Mac (you will have to build the project yourself in Godot, which is free, as you can't build for MacOS without a Mac and Youwin doesn't have a Mac).

I'm personally planning on integrating this into my project openseeface-gd in order to send scene data to OBS while having the controls open.

You can take a look at the codebase and see how the data is received and then applied

How long until some geeky girl in the free software community gives her a try and hooks her to the one, the only VTuber capture tool based on Godot? https://github.com/you-win/openseeface-gd

I've been working on OpenSeeFace GD for a few months now, and I can finally say I'm nearing completion in terms of usability.

I think thats just the version of the addon ? I think it just that blender version is too new? Because it says the addon was released in 2022. addon

I created a base mesh in the 3D program VROID before porting it to Blender via the VRM Blender add-on (link). I refined it to be the mesh I wanted within Blender itself (no sculpting, just pure box-modeling). After experimenting with various texturing methods, I ended up using Inkscape, a FOSS vectoring program to create the character textures. From there, I used the Cycles Toonkit addon by Cogumelo Softworks to get that cel-shaded look you can see in the Blender renders. Rigging was done with another Blender addon called Auto-Rig Pro.

I reccomend using https://github.com/saturday06/VRM-Addon-for-Blender/releases This Addon lets you export directly for VRM and manipulate Mtoon shaders. And modify blendshapes, properties, etc. only problem is that adding iFM blendshapes are buggy, they don’t seem to register unless you import it into uniVRM and re-export. Although they might’ve been fixed by now.

Vrm importing addon :https://github.com/saturday06/VRM_Addon_for_Blender

Aah thats a different pipeline yes , i am not familiar with any tutorials for this specific application but generally i'd first try to export it with https://github.com/saturday06/VRM_Addon_for_Blender

I do believe you can stream directly from both unity and blender , which i do not know how to do yet but someone here might. But from a vrchat model you should be able to find addons to export to vrm such as https://github.com/saturday06/VRM_Addon_for_Blender for blender or univrm 0.66.0 for unity version 2019.4.31f1 https://github.com/vrm-c/UniVRM/releases?page=4 But not 100% sure if those models need any adjustments before they can be converted to vrm

My issue is quite simple, i imported a VRM file into blender (using this addon: https://github.com/saturday06/VRM_Addon_for_Blender ). The import works fine and the textures and everything work as well, but it seems like the lighting on the object is just set as it is.