WK 09: Creating PBR Maps

I’ve done a lot of research these past two weeks to see what the methods were to create my pbr maps for the wine vessel that we scanned last time. You can see my final result here and below that I will explain my process:

25K Tris Rendered in Marmoset

Reference Image of the Wine Jar Artifact

Process and Challenges

One of the first challenges I ran into was figuring out how to reincorporate the blue hue back in to the model so it looks accurate to the reference image. When I tried incorporating the spec map that was derived from the cross and parallel polarized images, it would bring back its shine but I still had a very gold artifact which isn’t true to real life:

I tried doing research online to see workflows from other people/studios, and many of them seemed to manually make their own maps using the data from the albedo map that was made from cross polarized images and spec map. I decided to try to manually make my own maps using substance painter to match the reference image. I adjusted color/levels of base color map to get it to look as close as it could to the reference photo.

Speaker Scan

I also was able to get a scan in of a speaker that we have here at the lab. It presented me with a couple of challenges because it has a lot of metallic parts to it. I wasn’t quite sure how to approach this at first so I did some research online and found a blog from Dale Utt, who is the owner of True Edge Archive. They have a lot of experience with creating 3D scans for museums and had a helpful blog going over their workflow/process to create the metallic map. Feel free to check it out: Blog

In the blog he goes over how he has to mask out the metallic parts on the object and sets it to black to generate his own metallic mask. I haven’t gotten to this part yet but here is how the speaker looks in Reality Scan currently: