Breakdown: How to Reinterpet an Original Concept of a Sea Dragon's Lodge
Léopold Herlaud shared the workflow behind the Sea Dragon's Lodge, explaining how he created modular pieces to organize the scene more easily and discussing how he achieved the subtle wind animation.
Introduction
Hi! My name is Léopold Herlaud, and I'm a 3D Environment Artist currently based in Montreal, Canada. I've always been passionate about movies, video games, and new technologies. I got my first computer at a very young age, and quickly became fascinated by digital creation and virtual worlds.
After high school, I decided to study 3D art and moved to Montreal because of its strong entertainment industry. Over the years, I have had the opportunity to work at studios such as MPC, Reel FX, and Mikros Animation on projects including Ant-Man and the Wasp: Quantumania, Pinocchio, Mufasa, Paw Patrol, and several other animation and VFX productions.
Most of my technical skills came from professional production experience, but also from personal projects and constant learning. As real-time technology evolved, I became increasingly interested in Unreal Engine, optimization workflows, procedural tools, and game environment creation.
I spend a lot of time studying new workflows, following industry innovations, and experimenting with different techniques to improve both the artistic and technical sides of my work.
The Sea Dragon's Lodge
Before starting The Sea Dragon's Lodge, most of my portfolio pieces were more focused on atmosphere and vegetation. I wanted to create a project that showcased stronger architectural workflows using modular assets, trim sheets, and optimization techniques while still keeping a cinematic and immersive atmosphere.
I'm a huge fan of games like God of War and environments that combine grounded realism with fantasy elements. When I discovered the original concept by Liulianban, I immediately felt inspired by its atmosphere, unique shapes, and storytelling potential. What made the project especially interesting was the challenge itself.
I wanted to stay faithful to the original concept while adapting it into a game-ready environment using modular workflows and optimized assets, even though the design featured many rounded and irregular shapes. For references, I relied heavily on real-world architecture, weathered wood, fishing villages, roof structures, vegetation, and realistic material studies.
Since I wanted the final result to feel more believable and grounded, those references were essential for maintaining scale, texture consistency, and material realism throughout the project.
Modeling
I started by setting up the project in Maya using Unreal Engine scale conventions 1x1 to keep the workflow consistent from the beginning. Since I wanted to remain close to the original concept, I also kept a very similar camera angle and composition.
The blockout phase was mainly focused on modularity and proportions. I created simple modular pieces first to test how the environment could be reconstructed efficiently. Some areas translated easily, but the rounded walls and roof structures quickly became more challenging from a technical perspective. For the walls, I used Unreal Engine's boolean workflows to preserve modularity while achieving more organic shapes.
For the roof, I created curves using the CV Curve Tool and deformed the meshes with Curve Warp, which allowed the roof tiles to follow the structure naturally while remaining modular and reusable.
Most assets followed a traditional workflow using Maya for modeling, ZBrush for sculpting when necessary, Substance 3D Painter for texturing, and Unreal Engine 5 for final integration. Some wooden beams were extremely optimized, with only a few tris while still maintaining a believable silhouette.
For vegetation, I used opacity-based foliage shaders combined with World Position Offset for subtle wind animation. Grass was optimized using PCG Runtime Grass GPU workflows to keep the environment performant in real-time.
One of the biggest time-saving techniques throughout the project was reusability. Instead of creating unique assets for every element, I focused on flexible modular pieces, trim sheets, tiling textures, decals, and vertex painting to create variation efficiently.
Retopology
Most of the modeling work was done in Maya, while ZBrush was mainly used for sculpting and adding surface detail when necessary. Retopology and UV work were also primarily handled in Maya.
Since the project followed a game-ready workflow, optimization was considered throughout the entire production process. I focused on keeping the topology clean and efficient while maintaining enough detail for close-up shots. All the wooden beams have the same texture.
For UVs, I paid close attention to texel density consistency across the environment. A large part of the project relied on trim sheets and tiling textures, which significantly reduced the need for unique UV layouts and helped optimize memory usage while maintaining visual consistency.
Texturing
For texturing, I mainly used Substance 3D Painter alongside trim sheets and tiling materials. My workflow focused on balancing visual quality with optimization by reserving unique textures only for hero assets while reusing materials whenever possible. One of the most important aspects was layering and variation. Even subtle roughness changes, edge wear, dirt buildup, and color variation can make surfaces feel much more believable.
The brick wall was probably the most challenging material in the project. Using real-world references, I spent a lot of time adding edge breakup, color variation, surface damage, and material contrast to avoid repetition and make the surface feel naturally aged. I also used vertex painting extensively inside Unreal Engine to blend materials and create smoother transitions between surfaces.
Assembling the Final Scene
I first assembled the modular structure inside Maya using temporary materials before importing everything into Unreal Engine 5 for final scene construction. Since I wanted to stay close to the original concept, I kept a similar camera angle and overall framing. I liked the idea of having a path leading toward the house because it naturally guides the viewer's eye through the environment and also helps the scene feel more playable from a level design perspective.
The house itself felt very maritime, so placing it near the ocean felt like a natural choice. I also added birds, props, vegetation, ropes, fabrics, and smaller storytelling details to make the environment feel more alive.
For detailed distribution, I combined procedural tools with manual adjustments. Elements like vines used procedural workflows, while trims, decals, hanging fabrics, and ornament details were manually placed to add richness and variation throughout the scene.
Lighting
For the lighting setup, I was heavily inspired by documentation and lighting studies from Arthur Tasquin, which helped me better understand realistic lighting values and exposure workflows inside Unreal Engine 5.
I wanted both a realistic daylight setup and a warm sunset mood inspired by the original concept. For the daytime version, I used a directional light around 110,000 lux combined with an EV value around 15 inside the Post Process Volume with Local Exposure enabled.
I then added a softer secondary directional light to help reveal the front facade of the house, which was otherwise too dark. To finish the setup, I added a few subtle point lights in darker areas, volumetric fog, adjusted the volumetric clouds, and refined the image using slight sharpening, contrast, and color grading adjustments.
The sunset version was actually simpler. I mainly adjusted the directional light angle and intensity to achieve a warmer atmosphere while keeping the same overall setup. Since the result already felt natural and cinematic, I decided to keep the lighting relatively simple and avoid overcomplicating the scene with heavy post-processing.
Conclusion
I started researching and preparing the project around December 2025 after finishing my contract at Mikros Animation. Production really started in January, and overall, the project took me around two months of active work, although not always full-time every day.
The biggest challenge was definitely the modeling and modular workflow. Since many shapes in the concept were rounded and irregular, I had to constantly think about how to keep the environment optimized and modular without losing the original design. This project reinforced how important planning is, especially during the reference gathering and blockout phases.
Spending more time thinking about workflows early on can save a huge amount of time later during production. For beginning artists, my biggest advice would be to choose projects carefully and really analyze them before starting production. Understanding how you are going to build something is just as important as the final visuals.
There are also so many incredible resources available today, tutorials, conferences, documentation, and breakdowns. Learning where to find the right information and how to adapt workflows efficiently can significantly improve both quality and production speed.