logo80lv
Articlesclick_arrow
Talentsclick_arrow
Events
Workshops
Aboutclick_arrow
profile_login
Log in
0
Save
Copy Link
Share

From Concept Art to UE5: Building a Stylized Catalonia-Inspired Town

Tai Le discussed the creation process behind the Stylized Catalonia Town, explaining how he made six initial concepts to define the direction of the environment and detailing how he made architecture kits in small, medium, and large sizes to create the building sections more easily.

Introduction

My name is Tai Le Dinh. I'm a Senior Concept Artist and 3D Environment Artist based in Barcelona. My background is mainly in concept art for games, but over the past year I have been expanding my skill set into real-time 3D environment art.

Stylized Catalonia Town is my full personal Unreal Engine 5 environment project. It was a big step for me because I wanted to take my concept art experience and apply it to a complete real-time environment, from initial concept art to final scenes in Unreal Engine 5.

Project Goal

The main goal was to challenge myself to build a larger environment by myself within a limited timeline, and to understand the full environment art process better. Professionally, I already had experience with concept art and game assets, but I wanted to connect those skills into a full environment workflow. I wanted to handle as much as possible myself: concept art, art blockout, buildings, foliage, materials, set dressing, lighting, and final presentation.

Since this was a portfolio project, I focused on making the full environment feel cohesive and visually strong, while keeping the workflow grounded in a game environment art pipeline and realistic for a solo project. I did not treat the project as a fully production-ready shipped level. It was built with game-art thinking and still ran above 60 FPS, but because of the large scope and limited timeline, I made some performance and production trade-offs along the way.

Deadline, Scope, and Decision-Making

From art blockout to final, the production took around 3.5 months of full-time work. Because the scope was large, fast decision-making became very important. I had to move quickly and constantly decide what was worth polishing, what could be reused, and what was already good enough for the final scene. I could not treat every asset like a hero prop.

Instead, I judged many decisions based on how much they improved the full environment. This was one of the biggest lessons of the project. Finishing a large environment is not only about making assets. It is also about controlling scope, making decisions quickly, and knowing where the time matters most.

Hybrid Workflow: From Key Concepts to 3D Design

At the beginning, I created six initial concept images to define the main direction of the town. These concepts covered the most important areas and helped me establish the visual language, mood, architecture, color palette, and overall feeling of the world. I did not try to concept every single corner of the environment or spend too much time cleaning up the concept art. Since I have both concept art and 3D skills, I wanted to use the concept stage efficiently. The goal was to create enough information to guide production, not to make fully polished illustrations for every area.

From those six concepts, together with real-world references, I could already see what kind of asset library I would need: buildings, towers, arches, roofs, foliage, smaller street props, and other environment pieces. Once I felt the main areas gave me enough direction, I moved into the level blockout and started building the environment in Unreal Engine.

After that point, I kept the design process flexible. Instead of going back to paint a clean concept for every change, I used quick paintovers only when needed and continued designing directly in 3D. Many parts of the final scene changed during blockout and production. Some ideas worked well in 2D but needed adjustment once I saw them in the level. Working directly in 3D helped me solve those problems faster and made the environment feel more natural as a real space.

Learning and Preparation

Before starting full production, I spent around a year learning 3D environment art in my free time. I took online courses, studied how game environments are built, and practiced different parts of the pipeline. I also learned a lot at work, especially from my colleague Anastasiya Osichkina, who guided me through my first game-ready asset and taught me many environment art techniques.

During the process, I also took Environment Art Mastery by Thiago Klafke, which helped me understand how to approach a larger environment: blockout, modular planning, building kits, scene organization, and working inside Unreal Engine.

Inspiration and Visual Direction

The environment is heavily inspired by architecture around Catalonia, especially Barcelona. Some of the references included Hospital de Sant Pau, Casa Batlló, Girona Cathedral, and the farmhouse roof inspired by Casa Teixidor in Girona, among many others. However, I had to simplify a lot of those ideas to make them work in a stylized modular environment.

Catalan architecture can be extremely detailed, so the challenge was not to copy everything directly, but to capture the feeling of the place while keeping the shapes readable and reusable. I wanted the town to feel warm, colorful, and slightly storybook-like, while still being grounded in real architectural references from the region.

Architecture Kit

Because of the large scope and deadline, I used a modular approach for windows, doors, stairs, roof pieces, and several building extension parts. Instead of making only small wall modules, I also created larger building kits in small, medium, and large sizes. This helped me move faster and create more complete building sections without spending too much time assembling everything piece by piece in the engine.

This workflow was very useful for a solo project. It gave me enough flexibility to create variation, while still keeping production manageable within the timeline. For the buildings, I used two UV channels to add leaks, dirt, and roof color gradients. I combined this with vertex painting to make repeated pieces feel more different from each other and reduce the modular look.

I used 2 UV channels to add the leak, dirt, and color gradient for the roof, then combined with vertex paint to make them look more difference to each other.

Materials

For the materials, I used a few different workflows depending on the asset. For simple tileable materials, such as square bricks or broken tile patterns, I worked directly in Substance 3D Designer. For more irregular stone tiles and trims, I sculpted the forms in ZBrush, baked them in Substance 3D Painter, and finalized them in Substance 3D Designer. I also designed some patterns and decals to make the environment more interesting and give the town more personality.

Since the environment has almost no metal, I found that the AO map did not help much and sometimes added too much noise in indirect lighting areas. Because of that, I decided to bake some of that information into the base color and save the texture slots for custom color masks, height, and roughness.

Foliage, Props, and Other Environment Assets

Foliage

Foliage is probably one of my favorite subjects to work with. For stylized trees, it was faster for me to assemble the canopy shapes directly in Blender instead of using dedicated vegetation software. I modified the canopy normals by transferring data from a sphere. While testing the shader, I found that high roughness caused some strange shading in the Fresnel area, so I lerped the original roughness map with a custom Fresnel roughness value around 0.3 to soften the result. The foliage was not only used as decoration. I used it to frame shots, soften the architecture, and add Mediterranean identity.

Props

For the props, I kept them a bit more neutral because they needed to be reused across the whole environment. Most of them used a trim-sheet workflow rather than a sculpt-and-bake pipeline. This helped me create a larger set of props faster while keeping the style consistent across different areas of the town.

Rocks

For the rocks and large cliffs/mountains, I used a combination of triplanar detail and RGB masks to add color variation. I also set a clear priority between foreground and background rocks. Rocks in the playable or close-up areas had more detail and more defined sculpting, while background cliffs were much simpler. This helped save time and also made the main areas stand out more.

Sky

For the cloud texture, I found a tutorial from Tyler Smith on how to use a real cloud image as a texture. It was really useful and played a big role in creating the dynamic and mood for the scene.

Key Areas of the Town

The Harbour

I treated the harbour as the first “beautiful corner” of the project. This area took around two months because I was still testing the workflow and creating assets that would later be used across the rest of the level. After this area was set up, the rest of the environment moved much faster. Once the workflow, materials, building logic, and asset library were established, I was able to finish the remaining parts of the project in around 1.5 months.

The Blue Tile Gateway

This gate leads the viewer deeper into the town, but it also gives a nice view back toward the harbour area. The blue tile became an important color accent and helped this entrance feel more memorable while still fitting the overall visual language.

The Church Area

The church area was inspired by Girona Cathedral, but I simplified the design heavily to fit the stylized direction of the project.

Dragon Plaza

The original concept art for the plaza was much more colorful, but once I added it into the scene, it felt too strong and pulled too much attention. Because of that, I changed the dragon design into a more monochrome blue color so it could blend better with the environment. This was a good example of how some ideas needed to change once they were placed into the full 3D scene.

The Grand Palace

When you first arrive at the harbour, you can see the Grand Palace far in the distance. I wanted it to create a sense of curiosity and make the viewer want to travel there while enjoying the scenery along the way. This area was designed to feel more impressive and elegant compared to the lower town, with stronger symmetry, cleaner shapes, and a more monumental feeling.

Olive Farm

The olive farm is probably my favorite area. This scene was mainly focused on foliage and creating a softer, more peaceful mood. The farmhouse was inspired by Casa Teixidor in Girona. I wanted this area to show a different side of the town — more rural, relaxed, and surrounded by nature.

Lighting and Final Presentation

The whole project uses an evening setup. I wanted to keep part of the strong sunlight to highlight certain areas of the level, while still making the scene dark enough to support additional man-made light sources. For the windows, I used a warm emissive texture to create a cozy feeling and help the buildings feel alive.

For individual key shots, Unreal Engine made it easy to adjust the lighting to frame the focal points. I treated lighting almost like a final concept art pass: controlling focus, contrast, mood, and readability. Lighting helped bring the different assets, materials, foliage, and set dressing together into one cohesive world.

Performance and Portfolio Trade-Offs

Since this was a large solo portfolio project, I made some performance and production trade-offs along the way. The environment still ran above 60 FPS in Unreal Engine 5, and I kept the workflow grounded in a game environment art pipeline, but I did not push every part to full production-ready standards. For a shipped game environment, I would spend more time on optimization.

For this project, I focused more on final visual quality, modular workflow, and real-time presentation. I think this was the right choice for the goal of the project. Since it was a portfolio piece, the most important thing was to show the overall environment direction, visual quality, and ability to finish a large scene. I wanted to be honest about the difference between a portfolio environment and a fully shipped production level.

Conclusion

Looking back, this project became more than just a portfolio piece for me. It was a way to connect the skills I already had as a concept artist with the 3D environment workflow I had been learning over the past year. Because the scope was large, I had to be very practical with my decisions. I could not fully concept every corner or polish every asset like a hero prop.

Instead, I focused on building a clear visual direction first, then used Unreal Engine and Blender to keep designing, testing, and adjusting the town as the project developed. Finishing the environment from art blockout to final in around 3.5 months taught me a lot about speed, scope control, and knowing where time matters most.

This project also confirmed something important for me: a strong environment is not only about individual assets. It is about how everything works together, the composition, rhythm, mood, color, lighting, and the way the viewer moves through the space.

For artists working on large personal projects, I would suggest keeping the goal clear and a realistic deadline. Planning is important, but many problems only become clear once the work is in engine. Try to define the hero areas early, simplify your references, and focus on making the whole world work together. 

You can find more of my work on ArtStation and connect with me on LinkedIn.

Tai Le, Concept Artist/Environment Artist/Illustrator

Interview conducted by Emma Collins

Built for the Game & Digital Art Industry
Get Our Media Kit

Comments

0

arrow
Type your comment here
Leave Comment
Built for the Game & Digital Art Industry
Get Our Media Kit

We need your consent

We use cookies on this website to make your browsing experience better. By using the site you agree to our use of cookies.Learn more