muni-tuber

the vtubing app made for municorn (me), but also for you!

configuration

muni-tuber is in rapid development right now, so everything here is very much subject to change. lots of change. for now, here's how you can tinker with the app:

place images in src/assets

place the following .png images in the src/assets folder. for now, each image needs to be the same size. eventually, we'll be able to configure individual placement for images of different sizes, so that'll be nice!

• "head base" images:
  • quiet.png: for when your character isn't speaking
  • half_speak.png: head with half-open mouth for speaking
  • full_speak.png: head with fully-open mouth for speaking
• eyes:
  • eyes_open.png: open eyes, default state
  • eyes_closed.png: closed eyes, for blinking

adjust speaking activation levels

right now, volume thresholds are configured as consts in the source code. in src/main.rs, change the values of HALF_SPEAK_THRESHOLD_DBFS and FULL_SPEAK_THRESHOLD_DBFS as needed.

running

ensure you have a Rust toolchain installed. this app has been tested with Rust 1.71.0 (stable) on NixOS Linux and swaywm.

first, clone this repo with git:

git clone https://codeberg.org/municorn/muni-tuber

dependencies

this app uses eframe and egui, so according to them, this app should run just compile and run just fine on Windows and macOS.

if you're on Linux, you will need to install some extra dependencies first. if you use nix, this repo provides a dev shell that you can use with nix develop. otherwise, see eframe's documentation for info on needed dependencies.

ready?

once you're all set, it should be as simple as running

cargo run

the app should use your default microphone as input. enjoy!!

if you run into problems, don't hesitate to open an issue here!