Calibration

Calibration is the process of measuring your neutral face and expression ranges, then normalizing the raw ARKit values so VParamix can drive parameters consistently across model rigger and vtuber.

For example, if CheekPuff range is 0 ~ 1 but your face can only reaches 0.8 at maximum, calibration remaps your measured range to 0 ~ 1, so full expression still produces full output.

Why do I need face calibration?

Face calibration helps solve the problem where face-tracking data cannot fully utilize a model's parameter range.

Every person's face is unique. Facial structure, mouth shape, eye movement, and expression intensity can vary significantly between users.

Without calibration, VParamix relies solely on raw tracking values. As a result, some tracking parameters may never reach their intended maximum values, causing expressions to appear weaker or less accurate.

Calibration creates a personalized profile based on your facial movements, allowing VParamix to make full use of the available parameter range and deliver more consistent tracking results.

How does calibration works?

• Tracks observed maximums value that could achieve from face-tracking data.
• Normalized the new output scale to 0 ~ 1.

Previous Solution

Riggers often need to adjust the min and max value of the input / output value directly on the VTube Studio.

However, these adjustments are not personalized and may produce different results for different users.