Week Eight

gracieszymanski04
Oct 30, 2024
7 min read

Updated: Nov 11, 2024

Daily logs arranged newest to oldest.

11/3/24

Today, I worked with our compositors to fix the color correction issue I previously talked about. The issue was still occurring in this week's early comp versions. This is what was being seen in the comp even though what I was working with was a much deeper color for the caramel, and less milky.

Alice was able to show me what was happening in the color correction comp process and how the caramel render was being lightened.

After talking with both Drew and Alice, they were able to change the color correction to make it more accurate to what I was seeing, but still have the caramel blend with the scene. This is what we ended with, far less milky but still looks good with the shots.

11/2/24

Today, I focused on working on Shot 5, our hero shot. One of the professor critiques from last week was to make sure that the shaders looked consistent between Shots 3/4 and Shot 5. All the shaders being used were the same but, I realized the issue was because of the scales of both files. Subsurface looks very different depending on the size of the object. Aaron's FX file had the pint huge while Min's FX files were all accurate to the real world. This is a pipeline issue that should have been recognized weeks ago but that was our mistake and we've learned for next time and going forward. We decided to scale down Aaron's file. Scaled down, this is what the render looked like:

I focused for a while on adjusting the subsurface levels and the displacement levels and noise values in order to get the scene looking right again. I also had to change the intensity of all the lights to get it looking the same.

Next, I focused on updating my original lighting, as well as fixing the pint shader in order to response to mentor feedback. The first small think I noticed was how the reflections were concentrated to a ball at the upper part of the pint.

I realized that the cyc background was cutting off where the reflections could go. To fix this I changed the visibility settings of the cyc background in order for it to not block the other reflections, in the reflection of the pint.

Another visibility issue I addressed was the way the pint logo was being reflected on the other side of the pint, causing you to see two logos in certain frames. To fix this, I turned off visibility in reflections of the pint.

Next, I focused on the mentor feedback about the black lines around the sides of the pint. I realized that part of the reason there were black lines that made the plastic look thick, was because the pint model itself had pretty thick walls modeled into it. I went into the geometry and deleted those inner walls. I then turned on thin walled, inside of my material that was set on the new geometry. Below is a comparison with the left being the original geometry, and the right being the new geometry with thin walled turned on.

I liked the way the reflections looked and how the plastic looked a lot thinner with the new geometry, so I stuck with that.

I then worked with Professor Gaynor to further fix the dark line issue. When giving her all the information, she helped me to realize the problem was the HDRI I was using.

I used this HDRI to have a neutral background in the scene but also to add to the mood of the shot. The problem was, all the darkness was being reflecting and causing those dark spots. I went back to our original HDRIs taken and used one that was not desaturated and did not have the exposure down and used that in my scene.

After doing this, I also decided to add a bounce card to my scene. This was to add back in a strong reflection that was not as prominent once the ingredients filled the jar. This is what I ended up with for the pint material and its reflections.

Responding to feedback, I also adjusted the lighting of the scene by upping the rim light value as Emily suggested, and softening the shadows that were in the scene. I worry a bit that because of these critiques, a lot of the drama and moodiness is now missing from this shot. Though, due to the overall aesthetic of the piece, that might be ok and actually better. I am interested to see what the mentor's say this week.

This is what I ended up with, also tweaking the background blue to try and as closely match our ending shot as possible. (not 100% rendered, has some noise)

10/31/24

Today, I focused on reworking Shot 3 and Shot 4. I started with Shot 3. For this shot I needed to focus on the mentor critique of trying to get that light to reflect on the cookies and make them less flat. I first took another look a the lighting. I adjusted the HDRI and keylight to my liking and then I went ahead and added a rim light to the scene. I also added color to the keylight to match the actual light that would have been in the room, and to match our color correction better. Before the light was just white and now it has a warmer tone to it.

I also played around with the the properties of the cookie shader quite a bit to get the effect I wanted. I tried many variations of the roughness and IOR of the cookie. I spent a long time tweaking each value until I ended somewhere I liked. It was hard for me to get a value that reflected enough but also still looked like a cookie. I also played around with making the actual shade of the cookie brighter, as seen below.

I didn't love how bright the cookies looked here, and felt that they then looked too close to the truffles, so I kept the darker color.

I also tweaked the caramel shader to really bring out those reddish and rich colors of the caramel. This is what I ended up with for this shot.

Another thing I did this week was changed the displacement map dimensions. Before, the texture map was 1k, so to optimize render time for the final renders, I lowered the dimensions of the map.

This is what I ended up with for Shot 3.

When comped in, I still was not happy with the results. I felt it still looked flat and was being lost in the background like the mentors said. I then asked our comp artist working on the shot, Drew, to shuffle out the specularity AOV that we rendered with and play around with the values. This was the result after the extra comp tweak.

Next, I focused on working on Shot 4. Last week, I was using a piece of geometry to try and have the blue light fade off. This was unneeded this week as our comp artist, Alice, was able to make the light fade away in comp in a much better way when we rendered out two versions of this shot. It also caused issues last week with the non blue light renders, so I first went in and completely deleted that geo.

After that, I went in and I added a rim light, separate from the blue lights, so that the cookies and truffles would be sure to pop up more and not be too flat. These are the two versions of the renders for Shot 4 with adjustments.

10/30/24

Today, I noticed a color correction issue that was occurring and talked to our compositor Alice about fixing the problem. Both Shot 3 and Shot 5 were being color corrected after they were rendered out and comped in. This caused the objects in the scene to look different than what I was looking at. Below are the images that made it into the sequence. The colors are completely off from what I was looking at in both shots.

I talked with Alice today and the issue is going to be resolved for next weeks comp. Shot 5 is not going to be color corrected at all and I will just be adjusting the blue color to match the ending shot more, and the Shot 3 color correction is going to occur to the clean plate, before the FX is comped in.

10/29/24

Today, we met with the mentors and got feedback. Our project is beginning to wrap up and we are really emphasizing the touches. We also had a sound cut in last week's version that the mentors enjoyed. Here are the comments the mentors made about our 85% progress:

Tweak the lighting on Shot 3, right now it seems too even and the the cookies seem too flat. The fluid parts look great but the solids could use some work.
Shot 4, more light on cookies. Try to recreate something like how the caramel is reflecting the light.
Shot 5, soften the shadows on the ground, possibly add a stronger rim light. Add some highlights on the jar, perhaps use a bounce card or up the reflectivity of the material.
Fix the motion blur on the liquid trail of Shot 3 - push the noise to the head and maybe increase the size.
Try and make the liquid even more organic in Shot 5 - match Shot 3 more.
Shot 2 aspect ratio feels off of the computer screen, meant to be an invisible comp so add more reflection on the screen or maybe a gradient.
Increase the reflection/refraction of the pint material to get rid of the black lines around the edges.

I enjoyed getting a lot of feedback this week on the lighting/shaders as since it has been such a long project, I feel like I was hitting a wall with tweaking and fixing things. I think I really needed the direct feedback this week to make more progress. Here is my personal task list for the week.

Personal task list for the week:

Lighting on Shot 3
Fix past week lighting issue of Shot 4 and tweak lighting
Adjust lighting in Shot 4
Adjust shaders in Shot 4 to be more consistent with Shot 3 (professor feedback)

And my pipeline deadline this week is to have shaders and lighting to Min and Aaron by Saturday, early afternoon.