Week Five

gracieszymanski04
Oct 10, 2024
6 min read

Updated: Nov 11, 2024

Daily logs arranged newest to oldest.

10/13/2024

Today, Aaron applied my revisions to his FX file and the compositors put it together. Here is this week's result.

On a negative note, Min and I ran into another pipeline issue when putting my shaders onto her FX file. When the shaders were copied into her file, the texture was off and so was the subsurface of the cookie.

When I tried to open this file on my computer, another problem occurred as seen below. This was with double checking the HDRI file path was correct.

When Min pasted her FX into my shader file, the displacement completely disappeared. This is an issue we will work on resolving this next week.

10/12/2024

I focused on our hero shot, Shot 5 today. As I talked about before, this shot is going from a freezer background to a cyc background. To accomplish this I used a key light, a rim light, and I used an HDRI we took to create reflections and act as a fill light. One of the big things I did while testing this shot was test what color the background should be. To fit more with the original freezer, I tried using an off white background. I also played around with a dark blue background to match the dark blue Talenti uses in their commercials.

Upon seeing them side by side, I like the darker blue much better. I think that the materials pop much better. I also think it fits in nicely with the other past Talenti commercials. My only worry is that it will not seamlessly blend into the story as the ingredients are supposed to be in the freezer. I will be interested to see how it looks in the sequence, and what the mentors think.

After settling on lighting, I continued working and tweaked my shaders. I focused on the plastic for a long time, adjusting the IOR and roughness to see what looks best and while getting professor feedback. I also played around the with the level of reflection projected from the HDRI in the rsdomelight settings. Below are a few of the settings I was testing.

I also began testing just looking at the pint with no ingredients in it. After I tested I ended on these setting for my plastic.

I then started working on tweaking my shaders for the ingredients. Making minor color adjustments as well as subsurface level adjustment went a long way this week since I had a starting point. All of the shaders still need to be revised to be better but the one I specifically want to focus on improving in the coming weeks is the chocolate cookie crumble shader. Right now, the geometry of these is bumpy due to a mountain node. I want to check in with Aaron and decide if we want to keep it a mountain node or turn it to displacement. I also want to get the reflections right as I am unhappy with them as of now.

Here are my current setting for the caramel shader, gelato, and cookie.

Another thing I accomplished today was tweaking the pint model we are using. I originally modeled this pint in Maya and brought it into Houdini. When looking at the model I was unhappy with how the hollow bottom was looking. I felt it looked too thick at the bottom of the model so I selected those faces and made the bottom thinner. I also beveled the edge loop on the inside. These were details I forgot to take into account when originally modeling because I forgot the pint would be clear plastic and these details would be noticeable.

Lastly, I also adjusted the lid animation of this shot so that if falls more properly in place where it is supposed to.

Here is the final still image result of these adjustments and work from today.

10/11/2024

After meeting with Professor Fowler, the displacement issue was solved. After lots of testing we came to the simple conclusion that displacement in RedShift just cannot be sized down with an object, even if it is in UV space. The displacement just has to be changed at the object level to match the proportion of the object. When I was creating the cookie shader, I was using a large cookie to test displacement. When that model got put into Min's FX, it was scaled down to .3 percent. Therefore the displacement also needs to be scaled down by around .3 at the object level. It ended up being just a small issue with Redshift that took a lot of testing and troubleshooting to figure out.

10/10/2024

Today, I focused on trying to resolve the pipeline issue we are having. Originally, the issue was that the texture were swimming after the shaders were copied into the FX file. After getting some help from Professor Fowler, she suggested copying the FX into the original shader file to see if the issue could be resolved. This did solve the texture swimming issue but now a new issue has come up. The displacement is getting wonky when placed on the cookie as you can see below.

Here are the things I have done to try and resolve the issue:

Placed a rest node before the material
Placed a rest node after the object merge to the FX
Changed everything to UV space
Remapped the UV right before the FX nodes
Changed the displacement to tangent space

The biggest discovery I have made is that the shader and displacement look fine right before the copy to points occurs in the FX nodes. I have talked with Min, the FX artist extensively as well and neither of us can resolve the issue. We have made an appointment tomorrow for professor help.

On a separate note, I also worked on smoothing out the camera that I had previously made yesterday. I am still not completely happy with it but I definitely see improvement and I am waiting to hear what Aaron from the FX team thinks as it will effect his sims.

10/9/2024

Today I worked on the Shot 5 camera. The mentors said in previous weeks that they wanted more time on this shot so I changed it from 5 seconds to 7 seconds. I also adjusted the angle so that you can see the ingredients go into the pint a bit. I also made sure to include couple seconds at the end to show the completed product as the mentos showed. The camera is still a work in progress though as I am having a little trouble really smoothing the key frames out.

10/8/2024

Today, we were given more mentor feedback. Here are the critiques they gave us that we will be focusing on going forward:

Edit is good, story is clear
Check the shot one CG background perspective, it still looks a bit off
Swirls need art directing as they fly into fridge
Hero shot FX pieces should not feel heavy, more appetizing, lighter swirl
The shape of the FX curve in shot 3, swoop wider, maybe under her chin, same pace or slower
What if fridge was a cyc, not a fridge to highlight the tasty desirable ingredients, consider tying it back to computer screen perhaps, beauty of the ingredients, env should not be distracting
Lessen the sparkles in shot 3
Frost? Could be interesting - (soft no)

These critiques specifically effect my work flow with the lighting as I will be redoing the shot 5 lighting to now not have a freezer background and to be more theatrical. I am more excited about this change as I can now be more creative with the lighting. I felt a little restricted by the freezer setting because in order to make it look like a freezer, it had to bland. I put together a cyc lighting reference collage.

Another thing my group worked on today was our pipeline. We have had trouble getting our footing due to shooting days and the hurricane but we can finally have a consistent schedule. We have decided to make a weekly pipeline schedule at the beginning of each week so there is no confusion. Here is this week's schedule:

WEEK 5 Pipeline:

Today - Friday: Drew - making footage into exr

Wednesday: Gracie camera to Aaron for Shot 5

Thursday: Get files sorted for FX and Lighting/ fix texture swimming issue

Saturday noon: Min: hand off rough FX to Gracie for lighting

Saturday evening:

Alice: Website, Shot 6 and 8, fix CG room camera
Drew: Start on shot 7
Gracie: Shaders and Lighting to FX

Sunday 12:

Aaron: Hand off hero shot to drew, EXRS for particles
Min: more revised FX (Cookie fracture and more slo mo, new curve path)
Drew: Shot 3 particles around head, and defocus/focus

Sunday 4pm: Everyone edit slide with you initials

SUNDAY 5PM: PPT VIDEO

We also finally established a naming convention that is going to be used when transferring files.

Naming Convention: SOPRANOS_SHOT#_LASTNAME_TEAM_V0#_EXTRA