LogInJaechan Gwon guided us through the European City project, sharing how he imagined an old European city using scanned assets and explored its atmosphere through color grading.
Introduction
Hello! My name is Jaechan Gwon, and I'm a Concept Artist specializing in creature, character, and hard-surface design, while currently expanding my focus into environment design. I began my journey working exclusively with 2D tools, but quickly transitioned into 3D after realizing how much it could accelerate my workflow and help me explore complex ideas more effectively. Much of my 3D knowledge comes from self-study, online tutorials, and carefully studying the workflows of artists I admire.
During college, I took a filmmaking course where I gained a strong foundation in cinematic techniques such as camera work, framing, and composition, as well as an appreciation for the power of color grading to shape mood and atmosphere. This experience deeply influenced how I approach visual storytelling, and I continue to apply these principles to create more cinematic and immersive concept art.
While I haven't yet worked on commissioned projects, I've built a diverse body of personal work that reflects my growth and exploration as an artist. You can find more of my work on my ArtStation page.
The European City Project
Before starting any personal project, I like to define clear goals for what I want to explore and improve. For the European City project, my objectives were twofold: first, to push myself further in experimenting with color variation during the Photoshop stage; and second, to integrate scanned 3D assets as a way to quickly construct a convincing cityscape without having to model every element from scratch.
I intentionally chose to focus on a European city with ornate, historic architecture rather than modern high-rises, since the decorative details present a greater challenge to recreate. To support this direction, I collected a broad range of references, not only architectural and compositional examples, but also photography that captured lighting and atmospheric qualities. My goal was not to replicate a specific real-world location, but to design a setting that conveyed mood, history, and narrative depth.
My main focus for this project was to capture the distinct character of an old European city: narrow winding streets, densely packed buildings, and architecture climbing up a steep hillside. To begin, I created a rough 2D sketch from photographic references to establish the overall composition. Once the balance felt right, I transitioned into 3D, blocking out the scene with simple cubes to test the spatial arrangement. If the blockout worked, I refined the sketch further and adjusted the 3D layout accordingly. After finalizing the composition, I replaced the cubes with the prepared assets.
Workflow
I usually begin by browsing Sketchfab for scanned models. Once I find a suitable one, I decimate it to reduce the polygon count while preserving the overall form. From there, I use the Boolean tool to break the model into modular pieces, giving me the flexibility to reconstruct and reuse it in ways that best fit the scene.
When it comes to textures, I modify the original albedo maps to generate roughness and bump maps, which add more surface variation and detail. I also adjust the color and values of the albedo map to better match the mood of the project. At that point, the scanned assets are ready for use.
For most of the custom assets, I typically do simple blockouts and use an Array modifier. However, for the curved architecture in this project, I combined the Array and Curve modifiers. I drew a curve to define the building's shape, then applied the Curve modifier to the arrayed assets. This method allowed me to create curved structures quickly and non-destructively: if I wanted to adjust the building's form, I simply edited the curve; if I needed more height, I just added another step in the Array modifier.
Since this project was focused on concept art rather than real-time optimization, I didn't spend time retopologizing assets. Instead, I used the Decimate modifier to simplify heavy models while keeping enough detail for rendering. For textures, I manually unwrapped and mapped the more important assets, but for most of the models I created myself, I relied on procedural materials with object mapping in Blender. This approach removed the need for UV unwrapping, sped up iteration, and made it easy to adjust materials on the fly.
Lighting
Lighting is one of the most important tools for storytelling, so I always start by researching references to set the mood. I often study photography and cinematic stills to guide my direction.
I usually begin by selecting an appropriate HDRI to establish realistic reflections and global light. From there, I set up the main light source, add subsidiary lights, and then introduce a gobo for additional detail and variation. I prefer to keep the setup as simple as possible, since in my experience, overly complex lighting often leads to unnatural results. I also frequently use volumetrics and fog planes to add depth and atmosphere. For the background, I use emissive photographic textures, which provide a quick and effective way to achieve realism and mood. Lastly, I fine-tune the scene in Blender's compositor, adjusting values and adding effects like chromatic aberration and bloom to enhance the final image.
For color grading, I handled post-processing in Photoshop. After a little retouching on the Blender raw render, I used Color Lookups and Camera Raw for color grading. Refining values and hues in a procedural way, experimenting until I reached the atmosphere I wanted. This stage was essential for amplifying the story and ensuring the final image carried the intended mood.
Conclusion
This project took around 40 hours in total, with nearly 16 of those hours spent preparing scanned assets. The biggest challenge was adding secondary shapes to the buildings, without proper planning, this step can become extremely time-consuming.
Through this process, I gained a deeper understanding of efficient asset reuse and also discovered how much I could push the final look during the Photoshop phase. At the same time, I learned the importance of establishing the right mood and color balance directly in Blender, instead of relying too heavily on post-processing.
The most enjoyable part for me was working with scanned assets. They carry an inherent level of detail that’s very difficult to replicate manually, which allowed me to achieve a convincing result early on. That gave me more freedom to focus on composition, lighting, and atmosphere, rather than worrying about whether the project would come together in the end.
For anyone looking to dive deeper into Blender for concept art, I highly recommend the tutorials by Jama Jurabaev, Danar Worya, and Piotr Krynski. Their workflows and techniques have been invaluable in shaping my own process. Thank you for taking the time to read about my process.