Audio2Face + Machinima CC rendering tests

Tech Breakthrough with our talking avatars use.
I just managed to bring A2F CC4 talk & emotions from A2F as cache export, and combined them in Machinima to full body animated cc4 character. I can even update the characters hair, clothes, body motion In machinima!
  • Possibility to render in 4K 60 frm with RTX-real time
  • Possibility to render in path traced!
  • We can now start to study how to use brand new audio2gesture in Machinima!

This gives our customers a possibility to enjoy real high frame rates when working. Also RTX realtime rendering to 4K resolution is fast:

If customer want, he can render out the same thing in cinematic quality:

Match looks

Next thing we can make RTX realtime / Path Traced to match the lights & detail:

…same scene with Cinematic Path Traced version:

Two Cinematic Path traced variations:

More speed

We still need to squeeze out every drop of speed, here is CC3 & Pekka custom settings:

  1. Minimum acceptable quality 3.72 3.7 sec/frm
  2. High quality 8.5 sec/frm
  3. High quality iRedering 4 x 3090 3.7 sec/frm

More resolution

4K Path traced high quality comparision

  1. CC3 local A6000 32 sec/frm
  2. CC4 local A6000 43 sec/frm
  3. High quality iRedering 4 x 3090 14.15 sec/frm

Perfect example

Finally a perfect example of our Light Rig with CC4 Camilla.
3K render 
Local A6000
90 sec/frm