WHEN THE SUN WENT OUT
(From Out 2025 to Nov 2025)
Single-player FPS - Windows
SUMMARY
My main goal with this project was to learn different aspects of the Unreal Engine in a practical way. For that, I chose a genre I am used to playing and love very much: A slow-paced tactical(ish) FPS, with hints of horror, mystery, and sci-fi.
The game can be downloaded and played for free here:
DOCUMENTATION: GAME DESIGN
Before even opening the Engine, I created a board on Miro and started planning and defining different aspects of the game I wanted to make. For the first part, I described my Goals, Constraints, Elements, and Gameplay Beats for it.
A) GOALS
-
Theme: Survival horror; Cold Mountains; Investigation & Mystery; Sci-Fi.
-
Length: Up to 10 minutes.
-
Learn different aspects of the Unreal Engine. E.g.: Blueprints, enemy AI, and the Level Sequencer.
-
Slow-paced and High-risk.
B) CONSTRAINTS
-
Simple enemy AI and navigation.
-
All implementation needs to be done by me with Blueprints (I started from the Third-person template found in-engine).
-
Acquire (Free) assets from FAB only for 3D models, animations, VFX, etc. Not for functionality.
C) ELEMENTS
-
First Person Shooter.
-
Firearms (Player starts with one, but gains access to a second later in the level).
-
Transformed Human enemies.
-
Aliens as the final enemies and boss.
-
Simple puzzles.
-
Text or Audio Logs with lore.
-
Slow and somewhat tactical combat.
-
No fancy character movement.
-
Cutscenes or set-pieces never remove control from the player.
-
Slow enemies with simple AI.
-
Simple checkpoint/saving system.
-
Linear progression and level design, with some level of exploration.
D) GAMEPLAY BEATS
-
Optional combat with the final boss.
-
Beat with limited visibility and high tension.
-
Gigantic sci-fi room.
-
Sci-fi gate to separate "realistic" beats from sci-fi ones.
-
Show the final boss trapped before combat.
Not all of these aspects are present in the final game, but most of them are (or at least some version of). As with any game project, things change (and should) as it progresses. As a Game Designer, it is important to hone the skill of deciding what to keep and what to cut.

The image above shows a simple method I used to ensure all main features and mechanics in the game would help bring it closer to the design pillars. For me, players must have a cohesive experience, where every aspect of the game is aligned with the overall goals.
The Miro board used to document all aspects of the game can be viewed here:
https://miro.com/app/board/uXjVJBrGPMo=/?share_link_id=923364064458
DOCUMENTATION: WORLD BUILDING
With the theme and main goals set, it was time for me to fuel my imagination and creativity for the steps ahead, and nothing better than a good world-building and background story to make that happen. The secret is to keep it simple, though.
ALERT: Spoilers ahead!
A) WORLD BUILDING
-
Decades ago, the Sun weakened and temperatures began to drop hard. Snow and Blizzards are everywhere. Days are incredibly short, with most of the world covered by endless nights.
-
To make things even harder, years ago, a plague spread, which made the human immune system collapse. Even a mild flu became highly contagious and lethal.
-
The human population dropped significantly and is now on the brink of extinction. To avoid the plague, it is not uncommon to either sacrifice or exile the ill. Even if only with a cough or sneeze.
-
Survivors prefer to look for isolated places and live in small groups.
B) THE FACILITY
-
The second part was added to an abandoned military station, built to research a mysterious alien called "Mother Angel". Scientists discovered that its blood could cure any illness and make humans immune to any disease. But for that, a subject would have to receive great quantities of alien blood.
-
The scientists then started cloning the alien. Thousands of clones were created in order to farm their blood.
-
For an unknown reason, the facility and the research were abandoned in a rush.
-
Survivors found the facility when looking for a place isolated in the mountains. There, they found shelter, electricity, and a huge stock of food. Even though they were curious, they never managed to open the gate that would lead to "Mother Angel" and its clones.
-
After some peaceful time, another group of survivors found the place, and a massive battle took place by the bridge. While the first survivors won the fight, many were wounded and weakened. It didn't take long before they started to become ill.
DOCUMENTATION: ENEMIES

With the design goals and backstory in hand, it was time to define the foundation for an important aspect of a shooter: Enemies. The challenge for this project was to keep enemy implementation and AI simple, but still be interesting and challenging for players.
The first thing I did was to define the types I'd like to have, and then give each archetype a more detailed description. An enemy always need to have a goal and a reason to be in the game.


I also wrote some notes on how different enemy types could be grouped, or "combos". They were a good starting point when placing enemies on encounters, but were changed quite a lot after playtesting.
DOCUMENTATION: BEATS, PACE, AND FLOW
Time to list all the main gameplay beats in the level, give it a nice flow and set the pace. Easier said than done.
When writing the gameplay beats, I basically separated the gameplay into areas (even though some areas might be played more than once). For each beat, I classified it by intensity. From Low and safe, to High and risky.

Pace chart - Intensity levels for each gameplay beat.
The goal was to keep players invested in the game, but never overwhelming them with never-ending conflicts or boring them with too much downtime.
The image below shows the description for every gameplay beat in the level. For each of them, I defined:
-
INTENSITY: Low if downtime or safe exploration. Medium when some risk is involved. High is usually for action and combat encounters, where the risk of death is present.
-
MOOD: The emotions and mood I am aiming for.
-
GOAL: The main goal for the gameplay beat. The beat needs a reason to exist. It needs to add something to the game.
-
DESCRIPTION: Basic and concise description of what should happen.


As with many other things, beats didn't survive unscathed. Some of them were changed, while others were cut completely. Sometimes, things play out differently in your mind than in-game. As a designer, it is important to identify those moments and act accordingly.
Examples:
-
I had implemented gameplay up to Beat 04 and noticed that I need a moment with more intensity there. But Beat 5 was intended to be a high-intensity beat already. I couldn't have the two close to each other. In the end, I decided to change Beat 5 to have less intensity (and cut the Banshee enemy with that) and ramped up the intensity with a big-scale combat in Beat 4.
-
By the end of the project, I noticed it was already taking longer than expected, and I needed to wrap it up as soon as possible. Beat 11 was planned to have a nice cutscene where the player hops into a helicopter and flies around the facility, watching the outcomes of its action be revealed. It was there for me to learn the Level Sequencer. But since I had the opportunity to work with it in other Beats, I decided to simplify the whole thing.
LEVEL DESIGN: PLANNING
From the get-go, I wanted the level to be at the top of a canyon, located in cold and high mountains. But even when knowing what I want, it is always beneficial to gather some references to, again, fuel my imagination and creativity.
For this project, I used Chat-GPT and Gemini to generate a couple of concept art images.
From that, I spent some time planning the overall layout for the whole level. It is a good starting point to have a notion of the scale and where things'll be (roughly) in-game. Labelling areas with the gameplay beats and intensity level helps even further to stablish the pace and flow.

LEVEL DESIGN: METRICS
With most of the planning done, I jumped into the Engine for the first time to get the Player's basic movement ready, so I could work on metrics. As defined by the design goals, I wanted a slow walk speed and not many movement options, such as jumping and crouching. Initially, the Player couldn't Sprint, but I felt it was necessary to traverse the environment without prolonging downtime.
With that out of the way, I used Unreal's built-in modelling mode to build modular pieces to dimension spaces, corridors, doors, windows, etc. As well as cover sizes, and how big spaces would need to be to fit different combat situations.
Eventually, that map became a Gym Metrics level where I could test everything before putting them in the "real level". A couple of things that I used this space to test and dimension:
-
How much time the player needs to cover specific distances, both walking and sprinting. This would help me to define distance and avoid too much downtime when spacing gameplay beats.
-
Optimal range for the player and enemies in combat.
-
Optimal distance between covers to make players feel less or more safe when moving between them.
-
A space to test every enemy in a combat situation, including the whole boss battle.

Metrics used in the game.
LEVEL DESIGN: BLOCKOUT
Using all the information gathered so far, from design goals, layout, and metrics, I could finally start blocking out the level in Unreal. At this first stage, I was only roughly placing everything around, with the main goal of getting a good scale to spaces and distances. Not much detail was added.
The buildings were only blocks, with a rough estimate of where doors and windows would be. There was no interior. Cover placements were only placeholders to indicate where combat would happen, but they would be updated later on.
The goal was to be quick.
Once I was happy with the scale of everything, time to start adding some details, mainly to the interiors. At this stage, I'd also start to add possible interactables like loot containers, log objects, etc.
Cover Placement
During combat, players will spend quite a long time hiding behind cover, looking for an optimal line of sight, while trying to avoid being flanked. Not having a crouch mechanic in the game added quite a challenge when placing cover, as low cover wouldn't offer real protection, and tall cover blocks the view entirely.
Cover placement was the last thing in Level Design, as I waited for enemies' AI behaviour to be finished to ensure a predictable, yet challenging experience for players. When placing them, I thought of creating "lanes" where players would feel compelled to follow.
It is important to reduce the number of sides a player can be attacked from when hiding behind cover. If they need to worry about too many, they will soon feel overwhelmed.

A commonly used technique adds one lane on each side, while leaving an open "death zone" in the centre.
For the first and second combat, I wanted to have a more open area so players don't have to think about too many new things at once.
Using mostly slow melee enemies helped give more time to players who just got into the game.
The second combat was, by a long shot, the most challenging to design and define cover placement.
In this encounter, players would face ranged enemies smart enough to use cover to protect themselves, as well as a sniper at the top of the building.
To add complexity to this encounter, it can be faced twice. Once with the player coming from the radio tower (Green lines) and once when leaving the bridge control room, where the sniper is originally placed (Red lines).
But it was a very rewarding encounter, especially with the possibility of facing it from different angles, changing the dynamic of it considerably.
For me, one of the highest points of the level.


The encounter with the human sniper is an optional path for players paying attention to the environment.
On it, players face a single, stronger sniper, placed at the end of the cave. Unlike the rest of the level, this combat happens in a very dark environment.
While the cave is very narrow, I wanted to give players more cover lane choices than usual, adding a subtle puzzle feeling to this encounter.
The last regular combat encounter in the game takes place with the player confined to a narrow pathway. The contaminated water damages the player, but not enemies.
In here, there is only one cover lane, but plenty of cover for enemies to use for their benefit.
While it is possible to defeat all enemies without leaving the first cover, most players will flow with the covers, in a zig-zag of shooting.

IMPLEMENTATION: BLUEPRINTS
As someone with a strong programming background and love for the technical side of game and level design, one of my goals with this project was to learn Blueprints and implement all the game's mechanics and features using them.
In the end, I was surprised by how easy and simple they are to work with and became one of my favourite ways to add functionality to games without fully jumping into code.
Here are a few things I worked with in this project:
-
Using the built-in Third-person template, I implemented a "true" (players can see their body) FPS character.
-
Implemented input using the Enhanced Input System, defining an Input Map and Actions.
-
Used Animation Blueprints for the player, enemies, and the boss.
-
Used components for a simple inventory when acquiring and using fuel cans and nuclear canisters.
-
Used Widget Blueprints for the HUD, Pause Menu, and Text/Dialogue UI.
-
Used Blueprint Interfaces to avoid unnecessary casting.
-
Used simple inheritance on enemies.
-
Enemy AI using Behaviour Tree, Blackboard, and EQS. Including a custom Generator with Blueprints for cover.
-
Used different Triggers to implement level functionality.
-
Plenty of usage of Event Dispatchers throughout the game.
-
A SoundManager Blueprint to manage background music and ambience.
-
Used Level Sequencer for different sequences in the game, including both endings.
-
Used Data Assets to store all information for Logs that can be found in the game.
And many more...
Beyond Blueprints, I also worked with Animation Montages, Materials, VFX using Niagara, Chaos for destructible objects, and Physical Materials for contextual footstep sounds.
CONCLUSION
Learning Unreal was something I wanted to do for a long time, and I am glad I finally did. Overall, things were much smoother than anticipated. Many things in the Engine were very intuitive and second-nature to me, and I am now strongly considering working with it for my commercial personal projects as well.
Engine apart, working on an FPS project is always a blast for me. And once more, I proved to myself how I enjoy implementing enemy AI, which was, by a long margin, one of the things that I spent the most time on in the project.
Here are more images of the final project:










































