How a model turns its eyes for the gaze channels of an expression, after the VRM 1.0 lookAt extension.
The gaze channels are the lookUp, lookDown, lookLeft and lookRight presets and the eight eyeLook* ARKit channels. "expression" keeps them on morph targets and expression sinks, as a model without this record does. "bone" turns the leftEye and rightEye bones instead and takes the channels off the morphs, because a morph cannot rotate an eyeball mesh without shrinking it at intermediate weights.
How a model turns its eyes for the gaze channels of an expression, after the VRM 1.0
lookAtextension.The gaze channels are the
lookUp,lookDown,lookLeftandlookRightpresets and the eighteyeLook*ARKit channels."expression"keeps them on morph targets and expression sinks, as a model without this record does."bone"turns theleftEyeandrightEyebones instead and takes the channels off the morphs, because a morph cannot rotate an eyeball mesh without shrinking it at intermediate weights.Reference: VRM 1.0 lookAt (https://github.com/vrm-c/vrm-specification/blob/master/specification/VRMC_vrm-1.0/lookAt.md).