WK 05: Color Correction Workflow and PBR Maps

Since my last blog I have figured out how to properly color correct my photos in lightroom classic and am still in the process of finding the best workflow to extract pbr maps from my scans. Below I’ll go over my process for color correcting images and what I’ve learned on extracting pbr maps.

Color Correction Workflow

Once I’ve taken my photos I upload them to Lightroom Classic before bringing them into Reality Scan. Here I can adjust the settings and set things like the white balance and proper exposure. This really made such a difference and seems to help the software pick up more features.

Here is an quick gist of my workflow:

1. First install the ColorChecker Camera Callibration plugin for Lightroom Classic. This plugin allows the user to create camera profiles. https://calibrite.com/us/software-downloads/

2. Import the photos and click on Export. Under X-rite Presets click on ColorChecker Camera Calibration and I setup the profile for my camera.

3. I Proceed to select my images and click on the Develop tab. Here you will want to click on the image that has your color checker. I then select my camera profile, set the white balance, and adjust myexposure if needed.

4. I then go ahead and export my photos as jpgs. I recommend choosing a name for them in a sequence. So for example: Piggy-1.jpg.

PBR Maps

Lately I’ve been trying to figure out a good method to extract the specular and roughness maps from my scans. So far this is my current workflow:

1. Take photos of the piggy bank cross polarized and non polarized, for each photo that is polarized there has to be an exact same angle of that picture non polarized. Here I make sure to take photos of the color checker with and without the filter. I typically will take two photos with the color checker at the start of each loop I do.

2. Color correct the images in Adobe Lightroom Classic.

3. Bring the images into Photoshop, this is where I start to derive my specular map. This process took such a long time and I know there has to be a better way. I would take the non polarized and polarized photos then subtract them, and make them black and white to get what I needed for the specular map.

The images I noticed were not perfect because the angle of the polarized image vs the non polarized was not perfectly aligned. This is why you can still see some white.

4. Bring the cross polarized images into Reality Scan and create a high and low poly model using the simplify tool. I would unwrap and texture the high poly model, and reproject it onto the low poly model to get a higher texture quality.

5. I then export the textures and mesh from the low poly.

6. At this point I want to repeat the same process for the specular images that I derived from photoshop. Instead of starting the process over I made sure the new specular images had the same names as the cross polarized images, and redirected the file path to the specular images. This worked great and I didn’t have to worry about realigning my photos. From here I was able to export my specular map.

7. I tried to get my roughness map by bringing my exported specular map into photoshop and inverted the image and adjusted the contrast a little. It’s not perfect so I’ll have to mess around with this a bit more.

Result

Because the process took so long to get my specular images, I currently just have one side of my piggy bank done. I uploaded my work in progress to sketch fab so hopefully it will be pretty easy to examine. If you aren’t able to see the model try using a different browser.

Test to see how the first pass of pbr maps look on the object.