Bryan Ray wrote:We also have a Shader node that can do some of the relighting in 2d space without even needing the camera,
Sounds interesting. I'll have to play around with that. I've only seen the shader node used for 3d objects and renders that already have normal and position passes.
Bryan Ray wrote:or you can use Displace3D and the world position vectors to smash a high-res plane into the shape of the 3d scene and light that however you like. And, of course, the fog bit was just a color corrector. Fusion does also have some actual fog tools that interact with either depth or world position.
I guess my original question should have been, how would you create a world position map from 2d video?
Sander de Regt wrote:I wonder if meshroom like software could deal with the pointclouds of Fusion to create stand-in geo.
That reminds me of this video https ://youtu. be/rYs5LH6gzvU at the 4:30 mark he starts creating geometry from the point cloud.