Sculpting and Texturing a Stylized 3D Environment with Autumn Vibes

Introduction

Hello everyone! My name is Liliane Bruijn, but everyone calls me Lili! I'm a Stylized Prop and Environment Artist who started freelancing in November 2025. Before I got into game art, I didn't know what kind of profession I wanted to work in. I always loved to draw and would often draw anime/cartoon characters when I was younger.  After graduating from high school, I went for a culinary degree. I was only looking around schools close to my hometown, and didn't know game art was a thing. 

During my culinary studies, I met someone who was studying game art at the Grafisch Lyceum in Utrecht, the Netherlands. I went into a deep dive, and something about it pulled me into this world. I applied for the Grafisch Lyceum after finishing my Culinary Degree, and dreamed of becoming an environment concept artist. The movie Avatar, made by James Cameron, really inspired me to get into environment design. I rewatched the movie at least 20 times, and the world of Pandora still gives me goosebumps. I wanted to be able to create worlds like that and let other people experience that same feeling. 

During my studies, I had to land an internship in my 3rd year. I was told I lacked the fundamentals to land an internship as a 2D Artist/Concept Artist, and that it was better to learn 3D instead. I dedicated myself to learning 3D and improving my skills so I could land an internship and graduate. I had taken 3D classes before, but I wasn't super serious about it. I thought I wouldn't need it as a 2D Concept Artist, but this mindset was not helpful. I spent half a year grinding 3D after school to get the fundamentals. I also had a bunch of help from friends who were really good at 3D already, and this helped me out a bunch. 

The combination of texturing, sculpting, and getting better at 3D really made me enjoy 3D modeling. There was also an incredible artist who really inspired me to improve and start building dioramas and environments, and that's Jasmin Habezai-Fekri. The combination of inspiring artists, getting into 3D modeling, and wanting to be an Environment Concept Artist was what caused me to work hard and improve my skills to become a 3D Environment Artist.

What inspires my work is often environments and places that feel alive and cozy. Places with a lot of foliage, movement, and vibrant colors really get me hyped up to start working on environments.

Evergrove

After finishing up my Unknown Island portfolio piece, I was contacted by M3DS Academy to see if I was interested in joining a mentorship/course with them. I was really excited and took on this mentorship for 10 weeks. In the first week, I gathered a bunch of concept art and discussed it with the instructors to see which would fit well in the scope. There were two concepts I really liked, and I chose the concept of QY.

The reason was that I loved the architecture of the buildings and the giant trees in the environment, but I mostly loved the color palette. I didn't have a lot of orange colors in my portfolio, and I figured this piece could add some different colors to my portfolio. I also really love autumn and the colors of the leaves during the season, so that also fits perfectly with my taste.

For this project, I really wanted to learn new things I wasn't familiar with before. One of them was RGB masking. I did hear about this before, but I never used it during my projects. I also wanted to improve my lighting skills. I often struggle with getting amazing lighting in my scenes, and it's usually built over time with my previous portfolio projects. I also wanted to make beautiful trees and canopies that match the concept art as well. 

Workflow

During this mentorship/course with M3DS, I worked a bit with a different workflow than I normally would use. Normally, I start building the blockout directly in the engine, but I often have to adjust the scale, composition, and camera a lot during this process. During the blockout phase, I followed one of the instruction videos, where I built the blockout directly in Blender itself and added colors to the objects to get a clearer view.

Once I had almost everything from the concept blocked out, I imported the whole blockout directly into Unreal Engine. I set up a camera that matched the camera angle of the concept and added a directional light to the scene. Because I added colored materials in Blender, I didn't need to create materials in Unreal Engine. This saved a lot of time creating custom placeholder materials for each colored model and adding them to the model.

Once the blockout was in, some things did have to change. The camera angle was adjusted at least three times during the production of the environment. The tree and house on the left of the blockout also got moved more to the left as well, to create a more open area for the main building. During the production, the background also felt very empty.

The concept art had this same emptiness, but it didn't work as well in 3D as it did in 2D, so I added some mountains in the background that would support the composition and fill the empty space. The placement of the mountains creates a guiding line towards the main building.

Sculpting

For sculpting, I often use a workflow where I mask a surface and move that area inwards or outwards. This adds the feeling of worn wood or other damage on different surfaces pretty nicely and fast. For the wood in the environment, I sculpted one trim sheet and applied it to all the wooden elements of the buildings. This saves a lot of time, because hand sculpting every piece of the building takes a lot of time, and also a lot of texture resources. Being able to re-use one texture set on as many surfaces as possible helps a lot with optimisation.

Smaller elements like the crates, barrels, and boats were custom sculpted with their own textures. They were small enough that they didn't need tileable textures. In the final render, they don't really stand out, but I wanted to create them with the idea of a player ever getting close enough to view them.

Texturing

For the biggest elements in the environment, I wanted to use as many tileables and trimsheets as possible to create the final results. Big models need a lot of texel density in order to get crisp details and not get a blurry mess. 

For the houses, I would use an average of four different materials and RGB masking to get the results I wanted. The RGB masking helped a lot with blending the edges of the model and adding some custom details to the mesh. I wanted to make materials I could re-use as much as possible, which saves time in production and is something that gets done in the industry as well.

For the trees, I would also use a tileable texture, together with RGB masking to get some breakup in the tileable texture.

For the white tree, I used Vertex painting to get some dark spots in the areas that matched the concept art. The reason I didn't use RGB masking for this was that I added the RGB masking in the last week of production, and the Vertex painting was a lot earlier in the process.

I could have replaced the Vertex paint with an RGB mask, but that felt like doing the work twice, while the first results looked great on their own. I also used a moss decal on the big white tree. I wanted to have control over the shape of the moss, which is why I used a decal with an alpha texture.

Foliage

For the flowers and falling leaves, I first drew out the shapes of the leaves and flowers in Krita and set up the layout for the texture. After that was done, I imported the texture into Blender and modelled the leaves and flower shapes, and placed them on top of the sketch layout. 

Afterwards, I imported the petals and leaves into ZBrush and started sculpting some details in the high-poly models. I didn't want to go overboard on the details, since the flowers would be close to the camera in the scene, and I didn't want them to distract from the main building. Once the sculpting was done, I imported a plane into Substance 3D Painter and added the high-poly leaves and another plane for the bake.

I set up layers for each leaf, so the hierarchy stays clear and readable. I started on the orange leaves and built my way up to the other leaves and flowers. I also added some gradient areas at the bottom of the texture for the stems of the flowers. I didn't bother coloring within the lines, because an alpha texture would be used for the flowers. 

Once the texturing was done, I started building the flowers in Blender. Of course, they didn't look amazing in Blender, so I directly imported them into Unreal Engine to see how they would look in the scene. I hand-placed a few of the flowers that were very close to the camera in the scene, and in other areas, I used the foliage paint tool in Unreal Engine to scatter grass and flowers on the landscape.

For the tree canopies, I built a shader that only uses an alpha for the leaves. 
The colors, gradients, and Subsurface Scattering colors are all controlled in the material. Also, the leaf movement is set up in the material and can be adjusted directly to fit your scene.

I really wanted to achieve a fluffy and dense look for the tree canopies, so I took a small plane, bent it a bit, and started building layers of leaves to recreate the fluffy look that matched the concept art. It was a bit of trial and error. It's important to import your mesh into the engine and see what it looks like. I had to move some planes around a few times, but I did get the results I wanted.

Scene Building

Since I already had the blockout in the scene, I started building polished 3D models for the environment and replaced the blockout meshes with the real thing. Since I created this project with an Academy, they wanted me to create a simple prop to see how much knowledge I had of the entire pipeline. This asset would also serve as a benchmark asset for the environment.

Before starting on the big production, it helps a lot to have everything planned out from the start. This gives you a lot of mental space, and it also reduces the overwhelming feeling of having to create so much art for the environment. Planning where to use trimsheets, tileables, and which assets need to be unique helps a lot with the speed of production.

For example, if you make a trimsheet for all the wooden elements, all those areas will be textured pretty fast, and it saves a lot of time trying to make everything unique with its own textures, sculpts, etc. Same with the tileables: you make 1 tileable for the trees, and you have a big area of your environment textured already, instead of working on each tree by hand, one by one. The plan doesn't have to be perfect, and it's fine if things change halfway, but it helps a lot.

I created a crate model and showed the whole process in between to my mentors. Once everything was okay, I got started on the wooden trimsheet for the environment. I like to work from big to small when it comes to environments. You easily get lost in the details if you focus too much on the smallest assets first. 

When I start planning my trimsheet, I look at the concept and check how wide the surfaces are in the environment. The concept has a lot of smaller beams and edges, and not very wide surfaces. The only areas that needed a wider wood trim were the scaffold area and the wider platforms in the main building. I also had a tiny bit of space left over, and since there was some rope used in the environment, I added it to this trim as well.

Lighting

The lighting sure was quite difficult to get right during this project. My first attempt was not great; the lighting felt very flat, and I had way too many lights in the scene that were unnecessary. My mentor at this time told me to completely redo the lighting in the scene, and I followed one of their course videos for a basic lighting setup. I tried to apply the information to the environment, since the lighting setup of the environment was quite different from the lighting scenario in the video. 

The lighting started looking a lot better, but it wasn't fully there yet. My mentor, Simon Zafirovski, and I tackled the final lighting pass and made adjustments to the final lighting. We also tackled the fog and post-processing as well, and added fake godrays. 

We tried to stick to one directional light and three spotlights in the environment. Of course, we added two more, which were pointed towards the same direction, to get that nice lighting on the roots of the center tree and some of the buildings. 

Once the lighting placement was done, we started adding light catchers in the environment with cubes, which wouldn't be rendered in the final renders. There were areas in the environment that got light, which was distracting to the main composition and didn't match the concept. Creating light catchers gives you some art direction for the lighting in the environment. Sometimes, you want certain elements in the shadow, even though the directional lighting puts light on those assets.

In the post-processing, the elements that did a lot of the heavy lifting were the bloom, vignette, color grading, and the use of LUTs. During the process, a lot of these things got tweaked and adjusted, together with the fog and the lighting, until we got the final results. I sure did learn a lot during this process. When I do the lighting, I don't often go very deep into the post-processing and fog at the same time.

Conclusion

For the whole environment, I had around 10 weeks to build it from scratch until finish. In the first 5 weeks of the mentorship, I also worked on the Beyond Extent Team Challenge and worked a bit on some freelance gigs in between as well. So, giving an exact estimation of how long it took is quite difficult; I would say a month to two months tops. I did learn a lot during my time with M3DS Academy.

I learned a lot about lighting, RGB masking, the Niagara system, how to use Gaea, and so much more. I'm really thankful for the help and support they provided for me during those weeks. They also helped me set up my Rookie entry as well, which was incredibly useful.

Beginner Advice

PBR Workflow

The advice I would give to beginners who want to become Environment Artists is to first master the PBR low-poly/high-poly/baking workflow. Knowing how to make high-quality assets will be helpful when you get into environments. This doesn't mean you need to have an extremely high level of skill in this area, but being familiar with the workflow helps a lot. You could always try to make very small environments while learning about asset production, but be careful not to spread your focus too thin.

Dioramas

I tried to build an environment in the engine before, when I was more of a beginner, but I felt completely lost in how to tackle everything and what the steps were. I would have a half-baked blockout and didn't know what the best move forward was. Building smaller-scale dioramas helps a lot with those issues. You start with a blockout, getting the models in, UV/high-poly/low-poly/baking, and setting up the lighting. Once you have built a few dioramas and are getting the hang of it, you can try to expand your diorama into a bigger environment. 

I did quite a few dioramas before. What helped me get the full pipeline right was doing a mentorship with an environment artist who knew the steps you needed to take. During this mentorship, I was told to only create one corner area of the environment and just focus on what's seen inside the camera angle. This helped me have an area to put all my focus on and to polish, instead of creating an incredibly huge environment and getting lost in the process or having a lower-quality environment.

I know not everyone has these resources or money to spend on a mentorship. I do highly recommend them, but I know not everyone can afford them. There are plenty of videos online that go through the pipeline, and joining Discord servers like Beyond Extent and the Dinusty Empire is a good place to ask for help and feedback during your process. Beyond Extent hosts weekly WIP Wednesdays where you can submit your work, ask a question, and receive feedback from Timothy Dries. The Dinusty server also hosts critique sessions once a month, I believe. You can post your work, and you get live feedback on Twitch from Jeremy Estrellado.

Trimsheets/Tileables/Modularity

Once you get the hang of creating stunning-looking environments, it will be important to dive into how to use trimsheets in the environment, in which areas you use tileable textures, and how to break up repetition in those textures. I got this advice last year at an event by another Environment Artist. They told me that if I want to work as an Environment Artist, I'll need to know how to use those techniques and how to be as optimal as possible with the things I create, how to save time during production, etc. Trimsheets, tileable textures, and modular assets are a must in the environment art skill kit.

Lili Brujin, Environment & Prop Artist

Interview conducted by Stephanie Almogabar