Upload a portrait image with a close-up of the face.
For the best results, use one of the recommended sizes: 1024x576px, 576x1024px, or 576x576px (you can adjust the size in the Image Resize node).

Upload an audio file. Check the audio duration in the Load Audio node and enter it in the Duration field of the SONIC_PreData node.

More tutorials on https://www.youtube.com/@pixaroma