28 months later

28 Months Later - A post apocalyptic gas station

28 Months Later

CONCEPT IDEAS

Zombie outbreak setting.

1- Underground metro station (London underground/Sovietic brutalism)

2- Abandoned school

3- Sport Stadium

4- Police Station / Prison

5- Museum / Library

6- Skyscraper floor

7- Market / Grocery Store

8- Cinema

9- City Street

10- Farm

11- Truck Stops and rest area (Motel)

12- Airport

13- Resort

14- Fuel Station

15- Mall

16- Cruise Ship/Boat

17-Restaurant

18-Church

19-University hall

20-Amusement Park

Zombie References (Movie/Games)

MOVIES

"Night of the Living Dead" (1968)

"Dawn of the Dead" (1978)

"Day of the Dead" (1985)

"Zombi 2" (1979)

"Shaun of the Dead" (2004)

"28 Days Later" (2002)

"The Walking Dead" (2010)

"World War Z" (2013)

"The Return of the Living Dead" (1985)

"Train to Busan" (2016)

"The Mist" (2007)

"Dawn of the Dead" (2004)

"Resident Evil" (2002)

"The Dead" (2010)

"Zombieland" (2009)

"I Am Legend" (2007)

GAMES

"Left 4 Dead" (2008)

"Resident Evil" series (1996-2022)

"Call of Duty: Zombies" series (2008-2022)

"Dead Rising" series (2006-2016)

"DayZ" (2013)

"State of Decay" series (2013-2018)

"The Last of Us" series (2013-2022)

"Dead Island" series (2011-2019)

"The Walking Dead" series (2012-2022)

"How to Survive" series (2013-2015)

"Dead Nation" (2010)

RESEARCH

The first week  was spent researching ideas for the environment to be created. There are some key points regarding the project: it must be an environment with a zombies outbreak setting, and it must necessarily step out of my comfort zone.

Furthermore, it is a setting that I like a lot and I am sure I will have a lot of fun creating it. So, why did I write "stepping out of my comfort zone"? Because the environment I want to create has characteristics that I want to explore as an environment artist. Initially, the first idea I had in mind was to create a London underground station, but after writing a list of possible environments I decided to choose the one where I have the opportunity to best explore two of my weaknesses that I want to overcome with this project: Substance Designer and vegetation.

I therefore chose to create a gas station in the Nevada desert, which I think will be useful in this sense. With Designer, I will be able to create different materials useful for the creation of this environment, such as the concrete of the floor, the asphalt of the road, the soil of the surrounding areas. Regarding vegetation, I could create cactus, various types of grass and bushes in the scene, and a few trees. It is, in my opinion, a good compromise as we are not talking about an excessively vegetated area but it is still present and I will be able to explore this workflow without taking up too much time in my project schedule.

Planning and goals

Before starting the development of this project, I had set myself a series of specific goals. The first of these was to deepen my skills in hard surface modeling. During the first term I explored different workflows and developed a pipeline that I used during this project.A second goal that I set for myself was to improve my skills in the use of ZBrush for hard surface modeling whit it. I think that Zbrush offers many opportunities to create intricate details and surfaces, and I was interested in learning how to fully exploit its potential to create more realistic and detailed models.A third goal that I set for myself concerned the use of Substance Designer for the creation of maps and materials. This software allows for the creation of high-quality procedural textures, and I knew that it would be essential to achieve a realistic and consistent look in my project. Designer is an exceptional software and its importance in the typical pipeline of a game development studio is unquestionable.In the previous years, I have used Designer on a few occasions, and for this project, my intention was to explore the potential of the software.A fourth goal was to use photogrammetry to acquire realistic details of some assets I wanted to create.

In recent years, photogrammetry has become an important aspect in the production of assets and materials, and I want to make it a significant skill in my workflow. Preferring realism over stylized visuals, I think it's an excellent skill to add. Several programs allow creating assets or materials using photogrammetry, including Substance Sampler, my choice, and it is becoming increasingly easier to transform real objects into 3D assets or materials.A fifth goal concerned the creation of a large environment that would be modular and easily expandable. I wanted to create an environment easy to build in UE, and that could be easily customized and adapted to the specific needs of a project. In conclusion, my goal was to create a visually impactful and high-quality environment, using my skills in hard surface modeling, the use of ZBrush, Substance Designer, and photogrammetry. I tried to organize the project in a modular and flexible way also.

Final Design Concept

28 Months Later

28 months later is a 3D environment scene set in the Nevada desert during a zombie apocalypse in the 1990s. The scene features a gas station that has been overrun by zombies, and the players must navigate the environment to collect resources and survive. After careful consideration and research, I have decided to set the scene in the 90s instead of the 50s for example, or 60s, 70s or 80s, as those time periods have been explored extensively in many other games and media. The 90s offer a unique and particular aesthetic and cultural backdrop that can be incorporated into the gas station setting, while still providing the opportunity to fit elements of the zombie apocalypse. Overall, I believe that the 90s setting has added depth and nuance to the gas station scene.

Scene Description:

The scene is set in the Nevada desert, with the gas station being the focal point. The gas station is a small, run-down building with a garage and a small convenience store. The environment is dusty and desolate, with the mountains and desert stretching out as far as the eye can see. Two of my main references for the lighting in this gas station scene set in the Nevada desert during a zombie apocalypse in the 90s were the movie "Tremors" and the Mojave section of the game "Parasite Eve 2". The lighting in "Tremors" captured the feeling of being in a remote desert location, with bright, unforgiving sunlight during the day. The color palette of the movie was also a key reference, with its warm, golden hues evoking a sense of the harsh and dusty environment. The movie's use of strong shadows and contrast was also an important factor, as it added to the tension and suspense of the story.

Similarly, the Mojave section of "Parasite Eve 2" also provided an excellent reference for the lighting in this scene. The game's use of light and shadow was masterful, with a strong emphasis on creating a sense of dread and unease in the player. The color palette was also important, with a lot of muted tones and desaturated colors, giving a sense of desolation and decay. The game's use of baked lighting also allowed for dramatic shifts in mood and atmosphere.

Overall, both "Tremors" and the Mojave section of "Parasite Eve 2" provided excellent references for the lighting in this gas station scene. They captured the harsh, unforgiving environment of the Nevada desert, and used light and shadow to create a sense of tension, suspense, and unease. By drawing on these references, I hope to have been able to create a scene that is visually striking, immersive, and true to the spirit of the zombie apocalypse genre. The gas station itself is in a state of disarray, with broken forniture, and debris littering the floor. The garage doors are partially destroyed, the convenience store has barricades on the windows because of the zombie locked inside. The scene is designed to create a sense of chaos and desperation, with signs of the zombie apocalypse all around. The zombies themselves are a key element of the scene, despite they are not visible you can hear them locked inside the grocery.

Hypothetical gameplay:

The gameplay in "28 Months later" is centered around survival and exploration. Players must navigate the environment, avoiding the zombies and collecting resources. In this hypothetical game project the gamer would start their journey in this gas station in Nevada as the first scene before continuing the journey. The sense of desolation and loneliness will become a constant. The gas station immediately sets the mood of the game, and it is just an initial point in the game where the atmosphere and mood prevail over the gameplay.

Collection of references

The initial phase, which involved researching references and inspiration, made me change my mind several times about the general look of the project. The overall idea remained to create a very sunny environment, but I admit that the thought of creating a night time environment crossed my mind a couple of times. My mood board grew exponentially throughout the project, starting from this

and ending with this

Collecting references has become an essential and fundamental step in my 3D projects over the past three years. Before university, my collection of references was always very limited, with few images to give me a general idea of the object I wanted to create, but in the end, I did not use them as I should have, resulting in noticeable inconsistencies in the final shape. During the last few years, I have learned more and more how collecting as many images as possible regarding an object to create is not only useful for a general coherence but also helps a lot in creativity during the creation process. During the creation of the gas pump, for example, some images had details that my main reference image did not have, which I then included in the final mesh.

References help me with details, proportions, textures, colors, and lighting of the subject, but also with aspects that are not strictly related to the shapes themselves, such as their context in the scene. Furthermore, references (especially those of service areas) have helped me establish a visual language and a style, ensuring that the project maintains a consistent look and feel. This is particularly important when working on large and complex projects, where consistency and attention to detail are crucial to creating immersive and believable environments.

Having a general idea of the final scene, thanks to references of various gas stations in Nevada, has helped me to always keep a precise appearance in mind.

Modelling phase

The modeling phase for this project was intense and long, but as it is one of my more mature skills, it was a phase without major setbacks. My workflow is now well-established, with mid-poly modeling and the use of chamfers for weighted normals. However, I also had the opportunity to expand this workflow with the use of booleans, which allowed me to speed up the creation of some assets.

The first asset I created was the gas pump. Even before the blockout and everything else, I wanted to create this asset to have something "finished" in the scene right away.

This asset remained the same for several weeks, but the more I looked at it, the less satisfied I was. It wasn't perfectly in line with the other assets I had created in the meantime, but I postponed the revision because, all in all, it was a decent asset. Eventually, the revision came, and it was better that way.

The need to review and modify assets during the developmen of a project has become a recurring aspect of my workflow. I have learned that going back and making changes when something does not feel right or when aspects of the project change is always better than settling for less. It is a continuous process of iteration and improvement. Modifications may be made for a variety of reasons, such as improving the quality of the work, ensuring that it is properly integrated into the overall project, or adapting it to changes in the project's purpose or requirements. This process of revision and improvement is essential to achieving a final product that not only satisfies me but also helps the project to be the best it can be.

In addition to technical improvements, revising assets can also offer opportunities for creative exploration and experimentation. I have found that by being open to incorporating new techniques or ideas into existing work, even if it requires revising and modifying previous assets, unexpected and innovative results can be achieved that enhance the overall project. However, it is also important to strike a balance between revising and modifying assets and progressing the project. Focusing too much on revising previous work can cause delays and stop the project progression. Therefore, it is important to approach revisions and modifications with a clear understanding of the overall project goals and timelines. In the case of the gas pump, the mesh revision occurred after a discussion with Pat. I was explaining to him that I always tend to have clean meshes in terms of topology and avoid wasting triangles to keep the mesh lightweight. He pointed out that my modeling methodology is very clean and that nowadays it is no longer a problem to dare with a denser topology, and that adding a hundred more triangles could be the right choice if they actually add something more to the mesh in terms of overall shape, narrative, or simply make the mesh look better. So, that same evening I started redoing the gas pump from scratch, and the next morning I had completed it and was very satisfied with the work done.

This is the revised gas pump:

The asset is more dense, there are more details consistent with the chosen era, and overall it is a better asset. Leaving aside this little temporal digression that occurred later in the project, let's go back to the present.

In my workflow, I tend to use mainly 3dSMax to create the scene and then import everything into Unreal. However, this time, I wanted to proceed differently, also because of one of the goals that I had set for myself: modularity. This time, I only created the blockout in Max, and as I proceeded with the creation of the meshes, I regularly imported them into UE4 to start building the scene. The blockout in Max was initially like this:

As I created the assets, I moved the corresponding blockout out of the general scene so that I had a general visual overview of what I still had to do. The modeling phase lasted several weeks, without any particular hiccups.

However, I used Zbrush less than I would have liked. Only a couple of assets required a sculpted model, and in this aspect, I experienced my first failure in the goals I had set for myself.I want to specialize in hard surface modeling, and Zbrush is a truly useful resource for this. Unfortunately, for this project, I wasn't able to use it as much as I wanted, not only to improve my hard surface skills but also to reach a higher level of proficiency with Zbrush itself.

The time available for each university project has always been hard to complete by many factors, such as work or family. I would have liked to dedicate a whole day or maybe two to follow some tutorials and spend all day in front of the computer to learn and practice new skills more often, but I don't have this luxury.Surely, from now on, when I will have more time to devote to personal projects, I can choose to improve skills that I think are necessary to make myself more appealing to the market.

However, during university projects, I always had the difficulty of choosing between spending time following a course/tutorial or working on assigned projects, risking ending up in "development hell" and not being able to deliver on time. Thus, I always preferred to improve my existing skills more and put some techniques or workflows to experiment on hold.At the end I have create more than 150 assets, of course, some were created by reusing parts of meshes already made, such as the damaged sign from the car, where I used the already-made sign and carefully moved, repositioned, deleted, modified elements, or the damaged service pump, or the pipes in the buildings, created using the pipes from the oil tank, and so on.

PHOTOGRAMMETRY

If I reported a failure in the project (not the only one), I have a factor of success that satisfies me a lot. One of the goals I had set for myself was to experiment with photogrammetry.

Thanks to Substance Sampler and one of its latest updates, I had the opportunity to experiment with photogrammetry quickly, efficiently, and not of low quality. I used Sampler to create my head (then zombified in Painter), the shoe, and many materials. Let's start with the head and shoe. Here are examples of datasets used and imported into Sampler:

But these are just some of the photos used, shown as an example. My folder with the datasets includes 942 photos, divided into 12 folders, taking up 10GB of hdd space.

I experimented with different angle

I positioned the shoe in strange ways

I waited for cloudy days, I used two different smartphones, a compact Canon SX700HS camera, but in the end, my trusty mirrorless Fuji X-t200 allowed me to take fantastic photos and ultimately imported everything into Sampler.

OK. The head was not only the most enjoyable part of this project, but also the most enjoyable since I started doing 3D graphics, but also the most exciting and satisfying.

As I have already written, I made many attempts, with the help of my wife who followed my directions on how to take the photos, I tried different angles and different lighting, as well as different cameras. But when Sampler finally processed the right and correct model, I started laughing. But it was with Painter, when I started to texture, that my laughter became really loud.

Retopology and Unreal import were problem-free. After all, photogrammetry is not a complicated process. The same goes for the shoe, but as I mentioned earlier, I used Sampler and photogrammetry for many materials.

The pebbles that are in front of the Gas Station:

initially I used this (part of) photo:

which has been taken in front of the  Bede Island building at De Montfort University to create this material:

But I was very dissatisfied, it was not a nice material.

So I decided to redo it and took a photo in my garden (not the best picture but I just needed part of it):

the result material was what I was looking for:

I used photogrammetry and Sampler for many other aspects of the project. In my daily life, I have developed a habit of taking photos to use as references and as a base for creating materials or meshes using photogrammetry.

I took this photo of a manhole while accompanying my daughters to the pool:

the mesh decal in UE:

As I mentioned earlier, while I initially intended to use Substance Designer more extensively in this project, I found myself gravitating towards my tried-and-true workflow of using 3ds Max, ZBrush, Substance Sampler, and Substance Painter. Although I recognize the value of Designer in the creation of materials for a 3D environmental artist, I simply find myself more comfortable with the tools I've been using for years. However, this realization made me reflect on the importance of finding a workflow that works for you and your individual creative process.

While it can be tempting to try every new tool or software that becomes available, it's important to remember that ultimately, what matters most is the end result. If you're more comfortable using a certain set of tools, and they allow you to produce high-quality work that meets your standards, then it's perfectly fine to stick with them, but not remaining anchored in the comfort zone and falling behind in the technological evolution that allows this industry to advance so quickly towards new artistic heights.

In my future projects, I do plan on making a concerted effort to incorporate Substance Designer more into my workflow. While I didn't use it as much as I initially intended to in this project, I recognize its potential and the benefits it can bring to my work. I'm hoping that with more free time and without the pressure of tight deadlines, I'll be able to more fully explore and experiment with Designer in my upcoming projects.One thing I've learned from this experience is that it's important to strike a balance between sticking with what works and being open to trying new things like I did with the photogrammetry. By embracing both the familiar and the unknown, I can continue to grow as an artist and push the boundaries of what I'm capable of creating. I'm excited to see what new tools and techniques I'll discover and incorporate into my workflow in the future, and how they'll help me to achieve my creative vision.This is the material made in Designer:

I wanted to start with a simple material, and immediately thought of the one for the ground surrounding the gas station. I tried to reproduce a dry terrain with cracks and more rocky parts, and the result in Designer was not bad at all, but when I imported everything into Unreal, unfortunately, the visual outcome was very poor. So in the end, I decided not to use it. As I mentioned, this was one of the goals I had set for myself, and using Designer turned out to be a significant failure for this project. However, I'm not worried. Now that I'll be working on personal projects without tight deadlines, I can dedicate a few weeks to creating materials solely with Designer to enhance my knowledge of this software.

MATERIAL CREATION PHASE

Photogrammetry was an exciting phase of the project, but as I've mentioned multiple times, I've developed a deep "love" for Painter over the years. Creating materials with Adobe's software has always been a pleasure, and my proficiency with the suite has consistently grown. I feel highly confident today, continuously learning new techniques and producing increasingly convincing materials.

While I won't delve into the specifics of each material, here's an overview of some of the creations:

In this project, I made extensive use of atlases, which allowed me not only to control texel density but also to save on textures for all the objects used in dressing the scene. Many objects are of small size, so I grouped them into a single mesh, allowing me to create the material in Painter and then import them individually into Unreal. It's a highly useful workflow for managing maps.

I also experimented with a new workflow: UDIMS.UDIM is a texture mapping technique that has gained popularity in recent years and is increasingly used in 3D modeling. It allows artists to work with multiple sets of textures within a single UV space. Typically, a single UV space is represented by coordinates ranging from 0 to 1. However, with UDIMs, the UV space expands to accommodate multiple sets, each with its own unique set of texture coordinates. By using UDIMs, I was able to allocate different sections of a model's surface to separate texture tiles, resulting in higher resolution and more detailed textures. This is particularly useful for large or complex objects where a single texture map may not provide sufficient resolution. Each UDIM tile can have its own texture map, allowing for increased texture resolution and better control over specific areas of the model.

UDIMs provided me with more flexibility in my texturing workflow, as I could paint or apply textures to individual tiles independently. This allowed for greater control over the level of detail and seamless blending between adjacent tiles. In short, UDIMs offer a powerful method for managing and optimizing textures, enabling artists to work with high-resolution textures and texture complex models more efficiently. I used this technique to create the material for the oil tank. I had always wanted to experiment with UDIMs but never had the chance until now. This asset presented the perfect opportunity as it is a large, complex asset with many details. Being a hero asset, I wanted to ensure it had the right level of detail, and UDIMs allowed me to achieve that.

Unfortunately, towards the end of the project, I noticed that when I started adding the final touches on my home computer, I had to reimport the textures when launching the UE4 project. After seeking help online (on Reddit, Twitter, and the official Epic forum), it seems there is a bug in UE4 that requires reimporting the maps when starting the project on a new machine. It's as if there's a temporary buffer where the maps are allocated but gets cleared when switching PCs. I created a readme.txt file in the project folder to explain the correct procedure in case any artifacts appear on the oil tank mesh. I also included the map folder so that they can be reimported.

Anyway, here the asset:

As you can see, the dedicated material ID for the pipes has multiple UV spaces. Instead of overlapping UV islands to save space in the UV space, I was able to have more flexibility in texturing the pipes and create variations in the oil, metal, and overall details of the mesh.

REUSE AND RECYCLING> PROFIT

Asset reuse is a common practice in game development. I remember a Twitter thread discussing this aspect, and one post from a 3D artist at Naughty Dog caught my attention. They mentioned taking a helicopter created for Uncharted 1, make some modify, and then using it in Uncharted 3. I often utilize this technique to save time and expedite certain modeling phases.

The reuse of meshes was a crucial part of the project in fact. For example, when I created the narrative element (the car accident scene) I only needed to take the main sign mesh, make modifications by moving, removing, or slightly altering elements without affecting the UV, and then integrate it into the scene.

The original one

variation

From the most obvious assets

to the subtle ones:

For example, after creating the electrical pole and its variations, I created the street light using both parts of the electrical pole and the lamp created for the sign.

Another example I used the pipes from the oil tank to create the pipes for the main structures

UNREAL ENGINE PHASE

As I mentioned earlier, for this project, I created the scene in UE right from the start to have immediate feedback in the engine regarding meshes, materials, and the overall visual appearance. I didn't encounter any major issues during the scene creation, except for the vegetation. Fortunately, I was able to partially resolve it with the help of Pat, who knew about a command line (which I directly inserted into the project's .ini file) that allowed me to limit the short distance for shadow drawing of the vegetation.

The problem I had was related to the shadow drawing distance:

I didn't completely resolve it, but as mentioned before, with Pat's help, I managed to limit it.

I used vertex painting for the main area and the road, creating a custom material to achieve a more realistic asphalt texture with variations in colors, depth, and all the imperfections one would expect from a road or paved area. Here is the material:

I also added tessellation to the material to enhance its depth.

I made extensive use of decals as well, both to create variations in the asphalt and, of course, to add imperfections, varying roughness in different parts of the scene, and, of course, blood. Here are some of the decals used:

All the decals were created using Painter:

During the development of the project in Unreal, I wanted to add more dynamism. My initial intentions were to include animations like the grocery store door slamming shut due to the zombies trapped inside, or even hands breaking through the wooden boards of the barricade. However, I ran out of time, and the only dynamism I could incorporate was the wind effect on the vegetation and banners, along with ambient sounds that, despite everything, managed to make the scene feel more dynamic (at least to the ear).

Although I wasn't able to include those animations, I did create a small intro that plays when you start the project using screenshot, Audacity to edit the song and DaVinci Resolve to edit the video.

Here it is: