Week Four
- gracieszymanski04
- 9 hours ago
- 6 min read
Blog posts arranged newest to oldest.
To-do:
Refine flower vellum simulation
Get stem and leaves simulated
Refine flower and stamen simulations
Get the stamen to stop jittering and have no overlap
Add more noise to the flower petals and work on their movement
Work with Noelle on an animated "wetness" map for the petals, fix the petal UVs for subsurface
Refine shot 3 camera and ring animation
Create and refine frost simulation
Down the line - look into adding water droplets to shot 3
Mentor Feedback:
Shot 1 ending is too abrupt - work on the transition here - keep how the ring leaves shot 1 but change how it transitions to shot 3
Shot 3 - work on doing more AI concept for a better vibe
Textures and lighting on opening shot and end shot needs attention
Look into getting a motion media designer for the end title - this could make or break your whole project
Loved the shift back to the flower - more organic
The bubbles have progressed but look too thin
Work on getting a natural drop and bend for the flower in shot 3 - consider the weight of the ring
1/31/26
After prioritizing render tests today, here is the most recent render of shot 3:
First, I focused on assigning materials that Noelle created for the flower. She was able to make five difference petal designs, so I randomly assigned them to the flower.
With the textures on, I then sent a test frame to the SCAD render farm to make sure the USD pipeline was working.
The frame was a success! I then rendered out the whole shot on the farm for the first time.
As you can see from the video, I also have been again working with the camera and ring animation. Because the ring has to unnaturally pop out of the flower, instead of trying to hide it, I am going to push the idea of the ring being sentient. Therefore, I added in a little flip towards the end when the ring flies out.
I also was able to get the cryptomattes working thanks to Wren's awesome guide on how to use them in karma!
1/30/26
After working today, I was able to make a lot of progress and have some breakthroughs. Here is the most up to date shot 3 version:
First off, as shown in yesterday's entry, I was having some issues with getting the vellum to work properly. I also still needed to figure out how to add a collision object to the simulation. I was right in assuming that the problem was because of the thickness of the petals. Not only that, but the model of the petal needed to be simplified at the top. All of the different subdivisions was causing the crunch. I simplified the model, adding the ring as a collision, and got the vellum working!
*Note: I do still plan on adding thickness to the petals for rendering, but the node will just come after the simulation, not before.
Next, I continued working on using hair to simulate the stamen. I ran into an issue when actually pinning the hair to the animated center of the flower.
I'm still not sure what the solution to this is. For this week, I will be simulating the hair in a stagnant position and animating it afterward.
I then worked further on smoothing out the camera animation and the ring, as well as setting up the overall layout of the shot, including the flower bend animation.
1/29/26
One of the priorities today was getting a new camera working in the scene. Wren, our compositor, was able to send over a first draft of a camera yesterday. I spent some time today smoothing the ring animation out a bit and smoothing the new camera. There is a lot to still smooth out but here is the concept for the shot 3 new camera:
I also was able to start doing my first vellum tests of the full flower together. Below is my first test, but I see there is some primitives sticking together and the ends of the flowers are crunching up a bit. I will be troubleshooting this more tomorrow, but I believe it may be because I added thickness to the models (for the texture artist).
Lastly, today the group also successfully recruited a Motion Media Design student to help work on the project. After Cae reached out to her through a mutual connection in the class, Alessia has signed on to help with the end title, and to help us with adding text throughout the ad to help with the clarity of it. This means the group has now officially outsources two roles: motion media designer and sound designer.
1/28/26
After working today and starting the new shot 3, I was able to get a basic setup done so that Noelle, the lookdev artist on the team, can start texturing and possibly lighting the project:
To get here, I first began to research the types of flowers I could use for the shot. Most of the reference photos that I would appealing were peonies. I then found this articles explaining some of the types of peonies out there:
I've decided to use a single peony. I think the flower looks very appealing from the side, and it had a wide open center that the ring can fall into, even when it is not fully bloomed yet. That is key to the shot.
Reference images from Pinterest.com:
I not only looked for references online of the specific flower, but I also used flowers I found lying around and different weighted ring-like objects to take some real life reference videos. These will help me with getting a more natural look for flower movement, as the mentors noted:
I then began setting up the basic flower shape. I found a tutorial online that used a method of rigging the flower bud and bloom pose, copying the petals to points in a phyllotaxy, and then animating the petals' deform from bud to bloom. This was incredibly helpful as I have not rigged anything before, but this method worked flawlessly.
After modeling a basic petal shape, altering the phyllotaxy point positions and orientation, and modifying the petal bud and bloom rig, I was able to create this basic peony set up.
My first thought to get variation in the petals was to model different variation of petals. Therefore, using curves and booleans, I modeled these variations for the flower:
However, this method did not end up working. Still using the tutorial, I started testing the flower in the vellum simulation. When put into the vellum simulation, the modeled edges became crunched and did not work well with the sim.
Therefore, I am going to speak with the texture artist, Noelle, tomorrow, to see if we can create this variation through opacity maps in the textures, rather than through a model. I also will need to speak with her about the petals and their thickness. Right now, they have no thickness, but I know if she wants to use subsurface, they will need to have some sort of thickness for that to work.
The rest of today, I then also experimented with ways of simulating the stamen that the ring will land on inside the flower. I watched a few grass tutorials online and I think that using a hair simulation will be the best to simulate these stamen. Here is my first test with hair using a collision object:
Looking further into this, I also plan on setting the width of the hair to be similar in proportion to the stamen so that I can copy some spheres on the tops of each hair follicle (to get the stamen look) without causing overlapping geometry.
1/27/26
Today, we had another feedback session with the mentors. Here is the feedback we recieved:
Kyle:
Shot 1 ending is too abrupt - work on the transition here - likes the transition but it needs to fit into shot 3
Shot 3 - work on doing more AI concept for a better vibe
Shot 2 - be careful with the pop up color of the ring in shot
Textures and lighting on opening shot needs attention
Look into getting a motion media designer for the end title - this could make or break your whole project
Beck:
Loved the shift back to the flower - more organic
The shader on the end shot if feeling unrealistic
The bubbles have progressed but look too thin
Hailey:
Work on getting a natural drop and bend for the flower in shot 3
Consider the weight of the ring
Joleen:
Agreed there need to be a more natural drop and secondary motion in shot 3
Heather:
End shot needs some love
Keep working on the transition from shot 1 to 3
Weekly group to dos:
• Fx work on flower based on critique ( weight, opening if flower, physics)
• work on Camera Transistions (mainly 1-3 and ending shot)
• Background generations for the shots
• figure out render aovs and blurring issues
• update textures for new shot as well as materials for existing
• work on shot one animation
• find a motion media designer for the end card and text throughout
