Before working the shots, we need to decide on Agents.
We can provide generic pre-built agents, or we can receive custom assets from you and we build agents based on this.
If required we can also provide an example skeleton hierarchy, with which you can skin your own geometry.
Once the base character is created, then we need to add the behaviours.
These can be: custom motion capture, animations provided to us or from our library of pre-built actions
Next we will build up the agent to be able to perform all the actions needed for the shot, sequence and project.
Now we have our working agents, we can build up the shots
Typically we would work with your shot cameras and at least stripped down terrain geometry for placing the characters on
Next we will simulate the shots until they meet the required behaviour
Each iteration is provided to you for review.
Once the performance of the crowd is approved, we can deliver the shot back to you
This delivery can include:
Source scene file for any of the major crowd software packages
Rendered Frames (Beauty and compositing passes)