SCAD x The Mill: Week Five
Updated: Apr 26, 2021
Process Blog for SDGM 560, a 10 week collaboration with The Mill at SCAD.
Week Five Updates:
Now that look development is wrapping up, my main focus this week so far has shifted towards lighting. This is by far the largest environment I have ever lit, and the setting being nighttime lighting is definitely an added challenge. At Tuesday's session our discussion mainly focused on how much detail we were losing due to the dark lighting, and ways we could improve that. Different possibilities of time of day and light position changes were brought up, so we took the first half of this week to really explore the different options.
We ultimately decided to go with the top right HDRI, which is the original but set 10 minutes earlier in VUE so that we would be able to get a little more light in the environment. We did like a lot of the other iterations, however we realized the color temperature and light position would either take emphasis away from the effect or change the mood we were aiming for significantly.
After settling on the brighter HDRI we did another series of tests to confirm whether or not we wanted to change the light position, by rotating the original HDRI as well as the new one.
We realized that since there isn't a strong sense of directional light in this environment because it is dusk, that ultimately the reflections and maintaining detail would need to be our focus. Therefore we decided to not change our light position and stay with the backlighting.
After we settled on our lighting conditions I began shot lighting, focusing this week on only shot one and two. I worked on getting more reflections by adding cards closer to the car with the sky and clouds on them, as well as lights to emphasize the form of the car.
Shot 1: Original Iteration.
Shot 1: This weeks lighting changes.
Shot 2: Original Iteration (without composite).
Shot 2: This weeks lighting changes.
We also turned to working on the environment to retain more detail in our renders. By increasing displacement we realized we were able to create more surface detail for the light to catch. We also decided to add more small plants and foliage to the side of the road for the light to react with. We tried out working with paint FX within Maya, but just were not getting the results we were hoping for. The plants felt very wispy and really weren't adding the detail we wanted.
Billy then tried experimenting with cards to see if we could save on polygon counts and render time, but we still were not happy with the results. We ultimately went with actual geometry from Quixel, as we found it had minimal impact on render times.
Most recent pass (above) vs. First pass from last week (below).
This week our priority was Shot 1, and we will be working further on Shots 2 and 3 this coming week. We made a lot of great progress over midterms!