Post Production: Winter Quarter 4/4
- 3 days ago
- 5 min read
Updated: 1 day ago
Details work from class fifteen to class twenty(checkpoint class) of the Winter Quarter at SCAD. Blog posts arranged from newest to oldest.
Project Drive:
Project Shot List: https://docs.google.com/spreadsheets/d/1Y1-DKGdcjgM3FcTvOirFuHcpqTPOWDmjmlm5wzcjg6w/edit?usp=sharing
Other Group Member Blogs:
Hannah Kim | |
Benji Hopkins | |
Olivia Wesling | |
Sydney Relkin | |
Charlie Ragland | |
Mia Esparragoza |
End of winter quarter cut:
3/10/26
During group critique, we changed around some of the environment shots and notably combined 2F with 2G so that they better flowed together. With this change, I worked on the effect for this shot with the new camera.
Using the environment and the new camera, I drew a curve for the particles to follow so they will dance around trees and end where the magical bunny in shot 2G begins
I used two popnets for the effect, one with just spheres for particles and one with stars copied to points.
Similar to the sparkles in shot 2C, the particles are oriented towards the camera to ensure the stars are visible.
To create an interesting trail, the popnet has two pop source nodes and two popforce nodes. The pop sources are two spheres that follow the curve, with one trailing slightly behind.
Here is the look of the particles following the curve:
For the magical bunny, I am still waiting on the animation to be complete as there were some issues with needing a new rig etc. Therefore, I started the effect on the still bunny model I do have.
This effect is almost identical to the same effect on the base of the wings and on the vines of the tree in 2B with particles sticking to the geo and moving in a curved formation based on a gas.
Here are the effects combined in the shot with the new camera:
To do for 2F/2G:
Apply effect to the animated bunny once animation is complete
Work on Lookdev
3/8/26
For the tree effect, we realized that it would be better to use the actual tree that we scanned at the shoot rather than a random tree model from online.
I imported this new tree model and redid the simulation I had been working on in the last blog post, using the same vine method but refining the look to be cleaner.
I enjoyed how the special extra layer of particles looked on the wing simulation and also wanted to add that layer to this effect
I then also started playing with how this effect will dissipate in the background of shot 2B for continuity
To do for wings:
Refine timing of particle dissipation
Render out with proper camera for 2A and 2B
3/6/26 - 3/7/26
As mentioned in a previous blog post, I have been working on different layers of the wings and now just need to add it all together.
First I refined the base particles that cover the whole base and rendered those out
I then added another layer of particles in the same motion as the base layer that will be rendered with a larger pscale and different color to give more variation to the particles
To help with putting all of my layers together, I also rendered out the base geometry with just a red and green moving texture so that I will be able to easily shuffle the colors out in nuke and manipulate them fully in comp.
Taking all of these new layers and previous layers into comp, I was able to shuffle things out and use merge nodes to start piecing things together.
Here is the result from playing with these layers.
To do for wings:
further refine composite
add the trailing/falling sparkles
render out the wings with the proper animation and track for each shot
3/2/26
Cocoon Progress:
used vellum to add movement to the cocoon body to appear like someone is inside.
Gave leaves for the cocoon to sit on using vellum to also add more motion to the shot
Made sure to use the ground as a collision object for a more realistic look
To do for cocoon:
re-simulate hand pushing through cocoon effect with new plate
build onto the cocoon environment so that it is not just sitting alone
2/25/26
Today, the group took a good amount of time practicing group critique. I assigned all the members of the group to come into class with critique for all of the shots so we could all go over it and improve the project.
https://docs.google.com/spreadsheets/d/1nUMEiSIiiszvkMT7LILsRyr4zXfk1GTFHlerVpRe2Ho/edit?usp=sharing
Biggest changes:
The environment shots will be reordered - one before scene 2 starts, one will merge with 2G
1D is being cut
Lighting needs to be less flat, same with FX
The moon trail will follow the fairy more subtly
The mome design of the title will need to change colors
Blue spill issues will need to be resolved
2/23/26 - 2/24/26
One of the biggest issues that came about when doing the first full pipeline test with the environment and lighting artist, was the problem with the color space. The raw footage had been turned into ACEScg and therefore, for the composite, the renders also had to be in ACES but they were not rendering out correctly. I did some research and figured out the correct way to use ACES in the houdini viewport and how to export is properly with a custom AOV. I also typed up a report for my group members to use.
To further help with the rendering process, I then looked at the main environment file to try and help our environment artist with optimization.
Immediately I notes that all the tree and rock geometry was intersection with the heightfield and tried to reduce the geometry here. I set up the heightfield and other geometry to be booleaned.
Though this did reduce the overall poly count significantly, the time is took and the size of the cache file as it computed the boolean of the height field and the rest of the geometry was much bigger than just attempting to cache out more polygons on the first pass. This optimization attempt will not be used.
I did then move on to optimizing the height field for scene one and scene three. I created volumes from each of the cameras from these scenes and used them to delete geometry that is not visible by the cameras. This reduced scene one from 8 million to 3 million polygons and reduced scene three to 5 million polygons.
This file could be optimized much further if each shot was separated as its own section but with the way the lighting is currently set up in stage, shots are grouped together my scene, therefore, I do not want to change much more, but I am aware that if need be, reorganization can be done.
