Unreal Engine 5 MetaHuman Facial Animation from Song Audio File | Audio Driven Animation Workflow

Workflow of the previous movie Singing MetaHuman. The previous movie tested if Audio Driven Animation for MetaHuman Animator (Unreal Engine 5.5 Experimental feature) works with a song. Facial animation was generated from separated lead vocal of an audio file of the song and City Sample Crowds facial animation was mixed to improve the facial animation.

Original movie : https://youtu.be/DIpny2ao-pc

Unreal Engine version : 5.5.4
Editing and color grading : DaVinci Resolve 19.1

Music : Amazing Grace
https://artlist.io/royalty-free-music/song/amazing-grace/129079

Lead & back vocal splitter : LALAL.AI
https://www.lalal.ai/lead-back-vocals-remover/

Environment : Modular Concert Stage
https://www.fab.com/listings/b92c5261-9beb-44c2-8d2f-ff2214b54e2b

Body animation
Female Heels: Basics
https://www.fab.com/listings/4a9b6a4e-4068-41b7-85bb-f67b11632c6d

Facial animation : City Sample Crowds
https://www.fab.com/listings/903037e9-e1ac-4f41-96e8-1683c6fa7ad4

Cloth simulation : uDraper Real-Time Cloth Simulation Plugin and 3D Clothes Designer
https://udraper.com/
https://www.fab.com/listings/f94955ef-35d7-4d27-9f5f-8668e76cd77a