VirtualMotionCapture Protocol

First and foremost.

THIS IS NOT USING THE VIRTUALMOTIONCAPTURE SOFTWARE.

TTS Helper uses the VMC protocol. A lot of 3d (vtuber) software support the VMC protocol for receiving / sending data.

It also appears that if you use a VRM model, VRM version 1 models don’t work, but VRM version 0 models do.

But try it out and let us know anyway!

How to setup TTS Helper to send VMC data

Depending on the software you use, your LAN IP and port may be different.

Typically, the port will be between 39539 - 39541, but change it to whatever the software you use provides.

Testing the connection

Once you’ve decided your information is right, you want to test that TTS Helper is sending the blendshape data properly to your software.

When you push Test, TTS Helper will send random, silent, mouth shapes for 5 seconds.

If you do not see the models mouth moving, you may need to change the blendshape names.

Mouth Blendshapes.

Due to how TTS Helper generates mouth shape data, we only send “mouth open” and “mouth form”.

So it’s easier for us to just send A and E as the “open” and “form”.

However, if you have a model in 3d and another model to be the TTS avatar, the rigger for the model might not have left the mouth blendshape parameters as the default A and E.

Here, you’re able to change where TTS Helper sends data for A and E.

Does the mouth look too open or too closed?

If you’re having issues with the mouth blendshapes looking too small / large, we’ve added a slider for you to mess with to get the right feel.

100 means 100% of the value sent, which is the default. So if the mouth looks like it’s not opening enough, you may set it to 150 to make it larger.