Medieval Mill

Hi,

Here is my personal project for the Rookies Awards: a medieval mill. I will share with you my creation process, the various tools I used during production, what I specifically focused on, and the different problems I encountered.

I hope you will enjoy it!

Let's start with some renderings of my medieval mill in Unreal Engine 5 from different angles and with various lighting setups.

the rendering without textures to see how the light interacts with the various details of the scene,

this also allows me to check if the scene is visually pleasing without textures and if there are no lighting issues

                                      The modular kit that allowed me to create this scene

Now, various renderings of some assets that compose my scene. I rendered them separately to obtain a better visualization of my assets, but they remain within an Unreal Engine scene.

Next, we'll focus on my medieval box. The creation process is the same for each asset. We start by creating the low-poly model, which then goes into ZBrush to damage the edges of the box. I could have created the wood textures directly in ZBrush, but I wanted to concentrate solely on Substance Painter for the textures of my assets.

                       The different maps of my box exported from Substance Painter

                          A short video where I show my texturing process in Substance Painter

A wireframe of the various assets in my scene. I didn't really focus on optimizing my assets, whether it be for UVs or polygons. Instead, I concentrated more on the textures and how they would appear in Unreal.

Here is an atlas I created for my grass and ferns in Substance Designer. I followed a tutorial by Javier Perez to create my atlas and adapted certain parts of the process to achieve what I wanted.

Here are the different maps exported from Substance Designer that I will be able to use in SpeedTree next.

Now, in SpeedTree, I will shape my plants and grass to achieve a convincing result. Just like in Designer, I can create a random aspect to the placement of my grass, which will save me a lot of time, and I can have several different clumps of grass to avoid the repetitive look.

Next, we move on to creating the trees. This time, we'll do everything in SpeedTree. I'll start by creating a branch that I'll export to make maps, which will then be placed on a tree. With this method, I can create many different branches very quickly.

Next, since I learned how to create atlas in Substance Designer, I'll take advantage of that skill to create ivy leaves. Then, I'll make the ivy in SpeedTree. To add realism, I'll import the meshes that make up my medieval mill and place the ivy along those meshes.

Now, I will present to you the different graphs that allowed me to obtain the materials that compose my scene. This was the goal of my exercise: for several months, I learned to use Substance Designer through various tutorials, which allowed me to achieve these results. To save time, I used materials as a base and modified them to achieve what I desired.

                                        The different renders you see here were made in Marmoset.

Now we're moving on to Unreal. I've created a few short videos to show you what I've been able to do in Unreal for my project. For my path, I used Unreal's spline system to generate each map one after the other along the spline. I ensured that my texture looped seamlessly.

Here, I used the foliage system to quickly place my grass and trees. By adjusting various parameters, we can generate meshes quite easily with a satisfactory render.

I used the displacement system on planes to create my ground, which was also created in Designer. By overlaying the rocks with my ground, I quickly created detail.

Now, let me explain how I proceeded with my walls. I used a displacement map on a plane to generate details. My goal, once again, was not to optimize my scene. To achieve this result, I had to remesh my plane to add polygons. However, I'm aware that I could have used vertex painting to achieve a similar result, but I wanted to explore new horizons.

A short video to demonstrate what can be achieved with lighting.

And finally, to conclude, something I learned and implemented but didn't fully utilize is the POM (Parallax Occlusion Mapping). In my project, I didn't find a real way to use it optimally, but I keep in mind the various possibilities it offers for future work.

I thank you for going through my presentation. I hope it has been interesting to see my work process for this personal project, which I completed in about 2 months, taking into account my learning of materials.