%20copy_edited_edited.png)
%20copy.png)
Peanut "2.0"
- RB
- Mar 26
- 4 min read
“We can rebuild her. We have the technology!” Peanut 2.0 comes to life. As we learn from our past mistakes, we become better, smarter, and scrummier!
Sprint Goal
The goal of this sprint was to move deeper into production by strengthening our pipeline, improving the accuracy of our artifact (Peanut), and advancing environment and simulation work to support the narrative.
Peanut 2.0
This sprint focused on improving my initial reconstruction of Peanut. After inspecting the first version, and getting feedback from our Professor, I noticed that the albedo did not match the original lighting reference, which affected the overall realism.
Instead of trying to correct it through adjustments in post, I decided to rebuild the asset from the ground up as Peanut 2.0. This second iteration included updated textures, LODs, and turntables. By combining 50% Cross Polarized Diffuse (XPOL) and 50% Parallel Polarized Diffuse (PPOL), I was able to create a true albedo that is much closer to the original lighting references.
The biggest lesson here was that iteration leads to accuracy. As for someone who is still learning every small detail of the process, rebuilding with better understanding produced a stronger result than trying to fix issues too late in the pipeline. There will be more Peanut updates in the upcoming sprints as I want to get the most accurate roughness, metallic and specular maps. I am not only doing this for this one specific project. This is a process I want to master and apply to many different practical applications.
If you want to see the photogrammetry process and full lookdev, you can check it out [here]. This is a page I created to the full Peanut reconstruction,. This is more of a personal task for myself and I plan on updating that page to document the whole photogrammetry process from start to completion.
Water R&D and Simulation
During Spring Break, I used the extra time to focus on Water R&D for the environment. I researched both historical references and Unreal Engine workflows to develop a Shishi Odoshi (Japanese bamboo fountain) accurate to the Meiji era.
I built:
A modeled fountain asset based on reference
A Blueprint system to simulate the mechanical motion
A Niagara system to simulate flowing water
This process required a lot of iteration. Early simulations were too performance-heavy and caused instability. Through lots of testing, I was able to optimize the particle system to run more efficiently while still maintaining motion.
This was a valuable opportunity to explore a more technical area of the pipeline at my own pace, and it gave me a stronger understanding of how to balance realism with performance in Unreal Engine.
Collaboration and Team Support
Collaboration remained ongoing even during Spring Break. I worked closely with teammates across multiple areas:
Trained and set Michael up on the Metahuman workflow, including DNA files, asset migration, and parametric clothing.
Result: Metahuman is now fully rigged and animation ready.
Supported Blaise in environment construction, ensuring alignment with the animatic and historical accuracy.
Result: Gave the client a clear vision on what the environment could potentially look like.
Contributed ideas to the updated animatic, matte paintings, and workflows.
Helped build a client-facing PowerPoint for our work-in-progress presentation.
From a Scrum perspective, this sprint benefited from clear task ownership and knowledge sharing, which helped maintain momentum even as tasks became more complex.
Client Review: Inspection and Direction
Yesterday, we presented our work-in-progress to the client as part of our sprint review. Overall, the response was very positive, they were excited about how the project is shaping up and provided additional resources to help us improve historical accuracy.
At the same time, this meeting served as an important inspection point. The client asked key questions: What is the core idea? What should the audience take away? What is the thesis behind the story? These questions highlighted that while our visuals are strong, the narrative needs clearer intention and focus.
Backlog Adjustment
As a result, we are making a backlog shift for the next sprint. Instead of pushing forward into production, we are revisiting the animatic to refine the story. The goal is to advance the production on what is working while adjusting areas that lack clarity, ensuring that every scene supports a stronger and more cohesive narrative.
Challenges and Blockers
One challenge this sprint was navigating coordination during Spring Break, when communication naturally became less consistent. This highlighted how interconnected our workflows are, especially when certain tasks depend on others being completed first. It helped us identify areas where our process could be more flexible and better structured. Moving forward, we plan to adjust our workflow to reduce dependencies where possible and improve overall efficiency in future sprints.
On the technical side:
Initial HDRI lighting setups were not working as intended, requiring adjustment
Water simulations were too heavy before optimization
Texture limitations from RealityScan required rebuilding rather than tweaking
Each of these issues required iteration and adaptation, rather than quick fixes.
Sprint Retrospective: Key Takeaways
This sprint reinforced that iteration is part of the process not a setback. Rebuilding Peanut a second time led to a more accurate result than trying to fix issues late in the pipeline, and the same principle applied to other areas like water simulation and lighting.
Another key takeaway is that strong visuals are not enough without clear intent. Client feedback pushed us to think beyond execution and focus on the underlying message of the animation. This shift in perspective will guide how we approach the animatic and narrative moving forward.
From a workflow standpoint, this sprint also highlighted the importance of flexibility within a team pipeline. As tasks become more interconnected, improving communication and reducing dependencies will be essential for maintaining efficiency.
Adjustments for Next Sprint
Refine the animatic with a clearer narrative focus and defined thesis
Continue improving Peanut’s material accuracy (roughness, specular, metallic)
Optimize simulations further for performance and control
Improve workflow flexibility to reduce task dependencies
Overall, this sprint was about building on previous mistakes and turning them into stronger systems. The pipeline is becoming more stable, the assets more accurate, and the direction more focused going into the next iteration.
Comments