Creating Adventurer's Outfit Inspired by The Last of Us in 3D
Creative Cave told us about how they made an outfit for a male adventurer inspired by The Last of Us, explaining the process of modeling wrinkles and adding details to make it more realistic.
Introduction
Creative Cave is a creative production studio specializing in high-quality 3D art. We bring ideas to life through detailed design, strong artistic direction, and a commitment to technical excellence. Our team works across character design, real-time hair, grooming, and rigging, delivering visually compelling work for brands, game developers, and entertainment projects. At our core, we focus on creativity, collaboration, and pushing visual boundaries to create impactful, production-ready content.
Male Outfit
We started this project as part of our portfolio expansion. For the past few years, we’ve been focused on creating grooming for games, and now we’re looking to broaden our services into full character creation. This first asset is a demonstration of the quality and craftsmanship we can offer the market. We began with Joe’s original concept art from the popular game The Last of Us, developed by Naughty Dog, and added some personal touches to elevate the design in our own style.
Our reference board is always rooted in real-life materials, a mix of new clothing to accurately study pattern construction and garment fit, and worn clothing to understand how time affects fabrics and, more importantly, the color behavior of each material. In this specific case, we also included the different interpretations the character has had, both in the game itself and in the live-action series. Beyond being our main reference, these also served as a benchmark, as we wanted to demonstrate that the quality of our work can coexist within such a successful and acclaimed universe.
Our workflow takes advantage of the strengths of ZBrush and Marvelous Designer. We begin in ZBrush to create a blockout of the entire outfit, establishing the character’s overall silhouette. This blockout then serves as the foundation for building the garment patterns in Marvelous Designer.
Once there, we start the simulation process: adjusting the patterns and experimenting with the different materials available in the program’s library to achieve varied wrinkle weights. For us, this stage usually begins with the basics: ensuring that each garment fits properly on our base mesh. We then move on to internal cuts and seam lines, finishing with a combination of value adjustments for each cut to generate wrinkles of different sizes.
After completing this phase, we bring everything back into ZBrush, applying a quick retopology with ZRemesher to each garment so we can move forward with the high-poly process. This is where we finalize the work and take it to a more realistic sculpting level, fully leveraging everything we obtained from Marvelous Designer. We begin by refining the silhouette and primary forms, then polish and enhance the wrinkles in specific areas. The final touch is the most special part – adding story to each garment by working on memory folds, wear, and any damage or tearing that would come from real use.
This last stage – where we also break up flat areas – is done entirely by hand using a Standard Brush, always on a separate layer so we can control the intensity if we push things too far or need to remove something that doesn’t quite convince us. It’s perhaps the most important step, as it ultimately gives our outfits the realistic touch we aim for.
UVs
For this stage – both retopology and UVs – we handled everything in Maya. And while we approached this project as a game-ready asset, we didn’t want to create something fully optimized, so we allowed ourselves a bit more freedom with the polycount and the number of UV sets, without pushing it too far.
We always begin retopology by ensuring we have a clean quad-based mesh, and then move on to creating cuts around the entire silhouette using Multicut. Some UVs were brought in directly from Marvelous Designer, while others were created from scratch, depending on the garment and the approach we wanted to take.
For topology built from zero, we usually start with something simple: drawing large polygons with Quad Draw to cover as much area as possible and establish the overall topology flow, then using Smooth to generate additional polygons within that flow. We finish with support cuts using Multicut in areas where they’re needed – in this case, to preserve the shapes of the wrinkles overall.
Texturing
Texturing always begins with the fundamentals. We start by using materials from the Substance 3D Painter library to establish a base, and then gradually break things up and introduce visual disorder. This specific outfit belongs to a man living through an apocalypse, so the textures need to communicate that story.
After setting our base materials, we begin to subtly break down the colors by creating gradients across all the garments to add slight visual contrasts between each section. We also introduce subtle secondary hues to break up flat colors, generally through layers containing low-opacity tileable textures. These details are minimal but help create pieces that feel more visually engaging on a subconscious level.
Finally, we start adding dirt by focusing on the areas that would show the most wear. We break the dirt into different tones and intensities to achieve greater variety: from darker, almost black tones representing heavy grime, to softer, muted browns to simulate contact with soil, dust, and similar elements. Having all these color variations helps us create an asset that feels more believable and convincing to the viewer.
Lighting & Rendering
Our final renders were set up in Unreal Engine, where we used the Master Material we developed at the studio. This allows us to quickly connect and manage all texture maps across our assets.
For lighting, since we were presenting the asset against a neutral background, we focused solely on showcasing it in the best way possible. We used a main light positioned above the asset at medium intensity, complemented by two stronger rim lights on each side to clearly outline the silhouette. We then added smaller, softer fill lights to reduce overly dark areas and maintain visual balance.
Our post-production stage is a simple pass through Adobe Photoshop, where we run each render for final adjustments to levels, curves, and contrast. It’s a quick step, but a crucial one to enhance the renders and bring them to their full potential.
Conclusion
Building appealing assets is not easy, but it’s always rewarding to be able to admire the final results. Achieving this requires solid planning: from gathering references, which will serve as our guide throughout the entire project, helping us determine what we should and shouldn’t do at each stage, to making effective use of lighting to properly highlight every detail. It’s also important to emphasize that creating a strong visual impact relies heavily on contrast, both in sculpting and texturing. Nothing in reality is completely uniform – there are always nuances in color, variations in surface qualities, and subtle details that enrich the visual read.
The most challenging aspect of this project has been ourselves, as we set a high bar for the quality we want to deliver. We aim to provide our clients with assets of the highest caliber, so we push ourselves at every stage. But that doesn’t take away from how enjoyable the work is, or from the amount of learning it brings to each of our artists who contribute along the way.
For those who are just starting out, one of the best pieces of advice is to be highly observant of the world around you. If you want to create realistic work, nothing is more valuable than real-life references – and learning to identify the details and color variations that define them. It all comes down to paying attention.
