Week Three

Daily logs arranged newest to oldest.

Project 1 Directions: https://www.gracieszymanski.com/post/technical-direction-for-compositing-project-1-directions

Project 2 Directions: https://www.gracieszymanski.com/post/technical-direction-for-compositing-project-2-directions

1/27/25

Last class we began going over 2D and 3D tracking in Nuke for our project 2. Due to the winter storm in Savannah this past week, the class has had some bumps in the road when it comes to scheduling assignments. For homework, we were assigned to use a 2D tracker on our previous project and add a camera movement onto it so we could start familiarizing ourselves with the tool:

I also received critique from last class on project one that I want to note so that I may refine it in the future for my demo reel or for a resubmission:

• Try using a new AOV to individually grade the shadows on each ball to make them more cohesive.

---

1/23/2025

I have finished my project and here is the final product:

Here was my final nuke script:

In all I used an environment shadow layer, beauty layer, shadow layer, and occlusion layer. For AOVs I separated out key light and sky direct lights to use in my composite.

Citations for textures used:

---

1/22/2025

Today, I focused on trying to fix my issue that I spoke about in last week's blog where my viewport was showing one projection of my texture, and my render was showing another. I tried recreating the shader, flushing the caches multiple times, adding the elements to a new file, etc. After all this not working, I started going back to previous versions of my file to see when I started having issues. Everything was getting confusing so I decided to just remake my shadow projection in photoshop and take the opportunity to make it cleaner. I used shapes this time in photoshop instead of the paint tool.

After adjusting based on how far I needed to extend the shadow, this is the paint I ended up with for this assignment.

Even after completely repainting and starting over with this render layer, I was still having the same issue. I FINALLY figured out after all the troubleshooting that if I changed the fit fill from horizontal to fill, it fixed the issue.

After I could start progressing again after my big technical issue, I decided to rework my sphere animation. I have used Rigid Body Dynamics in Houdini before so I decided to give it a try in Maya. After some tutorials online, I added more spheres and created this simulation. This was also done so that I would meet the 25% screen space requirement for the project while still keeping to using marble spheres.

As talked about before, I also needed to revise my textures. I added a couple new textures to include: white marble, black marble, jade, and lapis lazuli. I didn't end up using the one crystal ball texture I originally cited before as well. I had to rotate these balls a specific way so that the pinch in the texture of the sphere was not visible in the render.

Next, I finally did my occlusion layer correctly after seeing the class lecture. We had the opportunity to do an object occlusion and a ground occlusion. Because my spheres were a little reflective in the shader, I didn't need to add an object occlusion because the bounce reflection was creating a realistic look already. For the ground occlusion, after analyzing the plates, I only needed a very small occlusion on the ground.

After all the renders are done, I will continue to tweak things in nuke until I get my final product.

---

1/20/2025

Today, a group of classmates and I met up at my apartment to take some photos for project two. These are the plates I got. I am not sure how happy I am with them and want to see what the professor says, but I am open to reshooting.