In order to perform a lip sync with a VRChat avatar, it is necessary to create an expression pattern that corresponds to the audio on the 3D model side.
Fortunately, the 3D model of the "window and Oko" used this time has an expression pattern of "Aiueo" equivalent to the vowel for lip sync, so use this.Since VRChat is made based on English, this material will be applied to English pronunciation.
Start Unity and open the project file set in the previous article, first select "Model" in "Hierarchy" on the left of the screen.
Change "Lip Sync" in "VRC_AVATOR_DEScriptor (Script)" on the Inspector on the right side of the screen to "Viseme Blend Shape".
Unityでプロジェクトを開いたら、Hierarchyの「Model」を選択する。VRC_Avator_Descriptor (Script)のLip Syncを「Viseme Blend Shape」に変更。The "Face Mesh" item added below Lip Sync is displayed. Click the round button on the right to open "Select SkinnedMeshRenderer".
Select Face in Select SkinnedMeshRenderer.After selecting, close Select SkinnedMeshRenderer.
Unityでプロジェクトを開いたら、Hierarchyの「Model」を選択。Select SkinnedMeshRendererで「face」を選択。Face Mesh will display 15 items called "Visame: ○○○".This is the setting item for linking the pronunciation and the expression pattern.
Each item is a drop -down list, and when you click it, select the expression pattern prepared for the 3D model.In addition, even if there is no applicable expression, "-none-" must be applied, but one of the expression patterns must be applied.
Face Meshの下に「Visame:○○○」という項目が15個表示される。発音と表情パターンを対応させるための項目だ。各項目はドロップダウンリストとなっており、3Dモデルの表情パターンから適当なものを割り当てることになる。Since the windowsille and this 3D model used this time do not have 15 patterns of expression, the expression pattern that seems to be close to each item will be set.Set the expression pattern suitable for each item while referring to the "PART 3: Visements and Other Blend Shapes" provided by VRChat.
Even if this item is set according to the guide, it will not be a perfect lip sync, so the choice of facial expression is rough and OK.Even with the following settings, the effect of the avatar is talking and the atmosphere is quite good.
今回適用した設定。発音を完璧に再現する口パクは難しいが、アバターがしゃべっている感じを演出するには十分な効果がある。