Creating a Game-Ready Cannon in an Arcane-Inspired Style
Arkady Nabirenkov broke down the workflow behind his Toxic Cannon project, explaining how he created an Arcane-inspired asset within a real-time game production pipeline and used the Diffuse map to achieve the show's signature style.
Introduction
Hi everyone! I'm Arkady Nabirenkov, a hand-painted Texture Artist. Drawing has been a part of my life for as long as I can remember. I never imagined it would take me this far — or that I'd stay on this path without looking back for a single day. My journey into the game industry began when I was eighteen and came to a simple realization: making a living by painting watercolor landscapes wasn't going to be easy. So I decided to leave my comfort zone behind and take a step toward something new.
After completing a few basic 3D art courses, I quickly broke into the industry as a 3D Artist, thanks in large part to my artistic background. My knowledge of composition, lighting, color theory, and traditional art fundamentals gave me a strong foundation and helped me get up to speed much faster.
What I love most are projects where artistic vision comes first — where you can clearly see the masters and artistic influences that inspired the team, and where the visual style has a strong, unmistakable identity.
A game's story or gameplay might not always be perfect, but I never miss the chance to admire a beautiful art style. To me, it's like walking through an art gallery, appreciating every piece for the passion and craftsmanship behind it. I genuinely enjoy working in 3D, but the closer a project is to something painterly, handcrafted, and truly unique, the more passionate I become about what I do.
Toxic Cannon
This project is currently my favorite and the best representation of everything I've learned throughout my career. It all started with a simple question: Could the visual style of Arcane be recreated within a real-time game production pipeline? More specifically, was it possible to achieve that look while working within the technical limitations of game-ready assets, where models have to be far more optimized than their cinematic counterparts?
My main goal wasn't to create a replica of an Arcane character, but to capture the same visual impression. I wanted viewers to look at the model and immediately think, "Yes, this feels like Arcane. I recognize that style." At the same time, every part of the asset had to be built using a practical game development pipeline, with all the optimization and production constraints that come with it.
References
I think gathering references ended up taking the longest. I actually went through Arcane frame by frame, capturing countless screenshots and carefully studying how the textures, lighting, and highlights worked together. I wanted to understand how the artists achieved such a distinctive visual style.
The more I analyzed the series, the more discoveries I made. It was incredible to see just how much artistic thought, craftsmanship, and attention to detail had gone into every single shot. The entire reference-gathering process only deepened my admiration for the project.
Every new insight into how the visuals were constructed helped me grow as an artist and sharpened my own understanding of stylized art. At some point, I started designing a weapon concept set in the Arcane universe. From the very beginning, I knew I wanted it to belong to Zaun.
I wanted it to feel like a functional device rather than just a weapon — something that looked handcrafted, improvised, and built from whatever parts were available instead of being factory-made. I settled on two core ideas that would define the design: a gas lighter and an industrial motor that powers the entire mechanism.
Modeling
When I looked at the model renders the artists had posted on ArtStation, I had a realization. The production pipeline didn't look like something that relied heavily on ZBrush. Instead, it felt like most of the hard-surface work had been done using subdivision modeling with support loops — and to my surprise, it seemed much more achievable than I had expected.
Even if my assumptions weren't completely accurate, this approach already produced a very similar visual feel to the original. So I got to work, building my model based on my concept. Throughout the modeling process, I tried to follow the design principles I had identified while studying the references — but without taking them too literally. My goal wasn't to recreate the exact cinematic workflow, but to develop a solution that made sense for a game production pipeline.
Here's the final high-poly model:
One of the biggest advantages of this workflow was that I didn't need to sculpt the model at all. I could use the high-poly mesh directly to bake normal, Ambient Occlusion, Curvature, and other Texture maps. That alone saved a significant amount of production time.
Texturing
This is probably going to be the main section of the article, so buckle up! The true star of this workflow is the Diffuse map. Seriously, once I broke down the texturing pipeline, I realized just how important the Diffuse map really is in this project. It carries far more responsibility than you might expect, and by the end of this section, you'll see exactly why.
But before we dive into the technical details, let's take a moment to appreciate the foundation of it all. At its core, it's simply beautiful, clean hand-painted artwork, built from structured, glaze-like brushstrokes with clearly defined shapes and a carefully controlled color palette.
Even though some areas may seem chaotic at first glance, there's actually a remarkable level of intention and organization behind every brushstroke. The following reference perfectly illustrates what I mean.
That said, I'll admit that I didn't follow this rule all the way through. I occasionally mixed both hard and soft brushstrokes to keep the final game-ready texture from becoming too visually noisy. I also allowed myself to use a slightly richer color palette than what I had observed in the references. While these choices weren't a perfect match for the original style, I believe they were the right compromises for a real-time game asset, and I'm happy with the final result.
I decided to texture the model in Substance 3D Painter because it's an excellent tool for PBR workflows. Although Arcane features a significant amount of hand-painted work — and I could have chosen 3DCoat for that reason — it's important to remember that the show's visual style is actually a masterful blend of PBR and hand-painted techniques. That's exactly why I felt Substance 3D Painter was the right choice for this project.
Another principle I set for myself was to avoid procedural shortcuts. I deliberately stayed away from generated materials, pattern fills, and filters that could have sped up the process. It was important to me that every brushstroke was painted by hand. I wanted the final result to feel like genuine craftsmanship rather than something assembled through automation.
Isolate. Paint. Isolate. Paint. Repeat… My workflow quickly became a simple mantra. Over and over again. My texturing approach was simple in principle, but incredibly meticulous in execution. I isolated every surface I wanted to work on into its own mask, allowing me to paint large, expressive brushstrokes that followed the form without worrying about affecting neighboring parts of the model.
This made the painting process much more intuitive and let me focus entirely on the shape and flow of each piece. Here's a quick example of what I mean:
One tip if you want to try this workflow yourself: straighten as many UV islands as possible. It makes selecting polygons for masking and painting much faster and far more convenient.
Another thing I paid close attention to was painting the bevels. Whenever possible, I made sure they received dedicated brushwork. Should you paint every single bevel? I'd say yes, more often than not. You're unlikely to regret giving them that extra attention — they play a huge role in defining the hand-painted look.
Color is another aspect I absolutely love experimenting with. I enjoy pushing color relationships and introducing subtle variations wherever I can. Sometimes I get a little carried away, but I think this time I managed to strike the right balance.
Roughness:
For a long time, I couldn't figure out how they managed to place highlights so beautifully — exactly where they were needed, while keeping other areas perfectly matte. I found it hard to believe that the artists were hand-painting the Roughness map from scratch the same way they painted the Diffuse map. I also doubted they were relying heavily on filters or grunge maps.
I had even seen other artists experimenting with painting brushstrokes directly into the Normal map to achieve a similar look, but none of those approaches felt quite right to me. Then, while studying one particular frame from the series, I noticed something that changed my perspective. The Roughness map matched the Diffuse map almost perfectly. That was my "aha!" moment. Suddenly, it all started to make sense, and I realized how the entire system could be working.
Isn't that amazing? Isn't it brilliant how such a beautiful result can come from a workflow that's both elegant and incredibly efficient? I honestly couldn't believe that something so simple could produce such a convincing visual effect. Sometimes the most effective solutions aren't the most complicated ones — they're the ones that make you wonder, "Why didn't I think of that sooner?"
Height Map:
Just when I thought I had finally figured out the roughness workflow, I noticed something else hiding in the textures.
Looking more closely, I realized I could see subtle layers of paint, exposed metal, and other surface details. What really caught my attention was that all of those height details lined up perfectly with the painted brushstrokes. "No way... are they reusing the Diffuse Map again?" Yes. They absolutely are. So I decided to do the same. Well... not exactly.
I used a slightly less detailed approach because I simply couldn't achieve the same level of precision. My asset uses a single texture set for the entire weapon, so I had to make compromises. Even so, reusing the painted information to drive the Height map gave me a result that stayed remarkably close to the visual language I was aiming for.
By this point, it becomes clear just how central the Diffuse map is to creating artwork in the Arcane style. It isn't just another texture — it's the foundation that drives almost everything else. Once the Diffuse map is strong, much of the rest of the workflow naturally falls into place. If you can get the diffuse right, you're already well on your way to achieving a convincing Arcane-inspired look.
Shader
With the textures finished, it was time to put everything together.
To build the final presentation, I treated Blender as if it were my game engine and started assembling the shader the same way I would for an in-game asset. (And for those who are curious, here's what the weapon looks like inside Unreal Engine 5.)
I'll briefly go over the shader setup I ended up with. I've also included detailed GIFs so you can recreate the same workflow yourself.
Here, I control the balance between the PBR shading and the unlit contribution:
This section controls the rim light:
And here I adjust the simulated front lighting. This turned out to be especially important for achieving artistically pleasing shadows around the rim light and creating a convincing shadow terminator:
It's a subtle detail, but I wanted the lighting to feel more believable and better refined in the small nuances.
This part handles the PBR contribution along with the AO intensity.
Overall, this shader replaces a significant amount of manual lighting work. I think it does a great job of making the model instantly more recognizable as something inspired by Arcane while keeping the presentation consistent under different lighting conditions. After putting all of these pieces together, this is the result:
One of the things I love most about the Arcane art style is the way metal reacts to light — how it catches highlights, shimmers, and shifts as the viewing angle changes. Recreating that feeling was one of my biggest goals throughout this project, and I put a lot of effort into getting as close to it as I could.
Conclusion
The production of this piece itself took me less than five working days. That said, those five days don't tell the whole story. Before I even started, I had already spent countless hours rewatching Arcane, studying it frame by frame, collecting references, and analyzing every artistic decision I could find.
On top of that, I had completed several previous projects in this style, which gave me the experience and confidence to tackle this one much more efficiently. If there's one thing this project reinforced for me, it's that good preparation dramatically speeds up the actual production process.
If there's one piece of advice I'd give to beginners, it's this: Learning how to study references — and developing the patience and observation skills to truly analyze them — is one of the most underrated superpowers an artist can have.
The better you become at seeing why something works instead of simply copying what you see, the faster you'll grow. Strong observation skills will teach you far more than any tutorial ever could.
Here are my ArtStation and LinkedIn.
Arkady Nabirenkov, Texture Artist
Interview conducted by Stephanie Almogabar