Building Fantasy Ancient Gateway In 3D

Introduction

Hello everyone, my name is Thanawat Mamak, and in this article, I will break down my thought process and workflows featuring a fantasy environment. I am currently a 3D art student at PlaygroundSquad in Sweden, preparing for my first role in the game industry. 

Before studying at PSQ, I was a self-taught 2D Artist who spent most of my time drawing, painting, and sketching a lot. What started as an attempt to improve my 2D backgrounds through 3D eventually led me into game jams, where people collaborated during a short period to make a game. I really enjoyed how each group approached ideas differently and collaborated to create something, even when the result was not perfect.

Getting Started

For my personal project, I knew I wanted a loose environment concept art that allowed me to fill in information myself. My main goals were to learn the material layering system in UE5 with RGB-masked textures, along with creating a larger modular environment.

I liked game franchises such as Devil May Cry and Bayonetta, and I wished there were more games like these. So I started searching for old, Victorian, and Gothic-style buildings while thinking of the worldbuilding and outfits they used in those games. My main resources came from Pinterest.

I saw this beautiful concept art by Cameron Mousavi that I liked a lot. I thought it would be amazing to play a game with this kind of beautiful architecture and scenery. I realised this concept art was a perfect fit for the aesthetics and my main goals. Hence why I chose to recreate it in 3D.

I created a simple daily task plan to track progress and maintain variation in my workflow. During the project, the plan helped me get an idea and estimate of what was left when I felt lost.

Along with that, I created a PureRef board to gather references in one spot. I gathered material references. architecture reference, similar buildings, natural weathering, and foliage. 

Modeling

After gathering what was necessary, I started with a simple blockout with a mannequin to define scale and composition. I took reference from this tutorial by Peyton Varney and this UE5 level blocking tutorial. My goal is to have everything there, even the lighting. I went with what felt like the general composition.

When the rough blockout was ready (Picture 1), I moved on to Blender to make parts that fit together, taking the concept art for reference. After the parts worked with grid snapping, I modeled it to a good enough state to tell what it is and export it into UE5. (Pictures 2 and 3).

After I had a decent representation and intended modularity of each part, my next step was to start deciding the best solution for each asset, which I will bring up next.

Workflows that saved me time:

• Mesh displacement into decimation. For the stone architectural elements, I used Substance 3D Designer, 3DCoat, and Blender. I displaced the arches inside 3DCoat with a heightmap I made in Substance 3D Designer and did some small manual touch-ups. After I decimated them, I baked and textured them in Substance 3D Painter.

• Skin Modifier. For the bench and the lamp post, I used a skin modifier with subdivision and bevel.

• Trimsheets. Thanks to trimsheets, I could easily change details on my arches and walls. Taking time to plan made it easier to imagine how each row would look and where it would be used. To make the trimsheet, I used Blender to model a simple trimsheet with cubes. Next, I sculpted medium details like stone-like edges in 3DCoat. Lastly, I used Substance 3D Painter to add more height details on the surface to give it more of a stone-like appearance.

Texturing

Once the base composition felt solid, I moved on to texturing, one of the most rewarding parts of the project for me. For texturing, I had to decide early on what to be uniquely baked, used with tiled texture, and where I can use trimsheets.

I had many ideas on what the materials could be when looking at the concept art. To narrow them down, I thought of the time period and aesthetics regularly and asked for feedback.

Here is a texturing breakdown for my arches, showing how the material works as well.

RGB mask layers:

Vertex painting:

Decals:

Edge decals:

Composition

In the beginning, I worked a lot with the main shot. Just like the concept. Eventually, I knew I had to make the surrounding areas more grounded and believable. Even more so when I was planning to create a cinematic with different angles.

This is where I had to think of designing the area, taking concept art as a reference, and also taking the liberty of my own ideas. My approach was to draw a rough 2D layout from a bird's-eye view, and then collect references that could strengthen my idea. 

Rendering & Lighting

I iterated heavily on lighting and sought feedback to refine the mood, while asking for feedback to get this result. I looked at many tutorials. Most notably, this lighting lesson by Feng always stood out in my mind, even before starting this project. 

I also found Karim Abou Shousha's YouTube channel extremely insightful.

Fill lights are important to bring even soft lighting to dark areas to make them more readable. They are there to mimic bounce light or ambient light. I used them here to brighten the cast shadows created by the 3D planes.

Key lights are important as well. The main purpose of them is to create a brighter tone on top of the fill light and shadow. A good key light should create an interesting shape and show clearer details of the subject(s) in your scene. For my scene, I used them to bring out the roundness of the tree canopies.

Fog was also used to create depth by manipulating contrast. I have a local fog volume on the left that mutes the background and at the same time defines the shop's roof. On the right of the shot, I used another local fog volume that mutes the transition of the greenery against the bright walls, creating a softer gradient. This is to shift the contrast to the center of the image. No fog cards were used in the project since I wanted a more real-time result.

Conclusion

This project took roughly eight weeks and two days to complete. The main challenges were matching the scale and spaciousness of the concept's foreground and finding the right camera angles. I spent a lot of time experimenting with focal lengths and compositions before landing on something that felt both faithful to the concept and visually strong in 3D.

My general advice for beginning artists is to consume media relevant to what you want to or will work with. Consistency matters more than talent. Start small, keep learning, and stay curious. The more you create, the stronger your visual library becomes. Start with small things you enjoy, and bigger goals on the side. You will always want a big goal and many small ones. 

If you feel like you are uncertain or intimidated about something, that means you need to research further and practice more deliberately. That was my motto during the project. 

Thank you for reading.

Thanawat Mamak, 3D Artist

Interview conducted by Emma Collins