Scroll down to see the stages of development. Scroll to end to view the result of the stage 1 deliverable.
In this stage we develop Molly’s neutral head scan. We start with a neutral scan and remove the hair.
Next the scan is cleaned. Hair and eyes are removed. High frequency skin detail is applied. The topology is optimized for facial animation.
High-res neutral head, cleaned.
Next, the texture work begins. The scanned texture is cleaned and we go though a process of “de-lighting” where specular highlights and shadow detail is removed. This is so that the avatar’s skin will react naturally to dynamic lighting in the realtime environment.
Specular maps are applied and tested separately.
Eyes and teeth are created as separate geometry and textured individually. Eyebrows and eyelashes are digitally “grown” based on reference photography.
Next, spec and gloss maps are checked with eyes, along with subsurface skin maps which are needed to produce realistic skin rendering. Note the light scattering in the ear when backlit.
For the hair, we brought in a hair model to create a style based on reference photography.
The hair model was then scanned and the result was merged with the Molly head.
Next, we grow CG hair strands that cover the scanned hair. This creates more realistic hair rendering.
When all the maps are in, it’s rendered in engine and and tested in a variety of lighting conditions.
Finally all the assets are rendered together in real-time using Unreal environment.
