Research and Development Project “MAGNI”
At Eighting, we embrace the slogan “Keep challenging ourselves”, actively exploring new technologies and methods of expression.
As part of this commitment, we continuously conduct R&D utilizing cutting-edge technologies and explore a wide range of genres and techniques.
Eighting has previously announced the results of its Unreal Engine 5 research and development project “RAIZIN” and its technology verification project for up to 64-player close-quarters battle royale “BRAVE”.
Building on these experiences, we continued our R&D efforts, and this time we are pleased to introduce our open-world technology verification project, “MAGNI.”
Exploring Open-World x Photorealism
In MAGNI, we challenged ourselves to achieve both an open-world environment and photorealistic visuals.
In Japan, There are still few examples of large-scale open-world title development, and it is extremely rare for a request to come in for an area of development where we have no prior experience.
That’s precisely why we felt it necessary to actually create one ourselves to identify potential problems and prepare solutions, ensuring we could handle such projects when they came.
Creating vast, natural landscapes with dynamic lighting, weather systems, and realistic environmental behavior—from terrain and vegetation to atmospheric changes—presents numerous technical hurdles.
In “MAGNI,” we tackled these challenges head-on, exploring how to build an open-world game efficiently with a small development team.
This project represents more than a technical experiment—it embodies Eighting’s commitment to breaking new ground.
Prototype Development and Key Findings
Now, let me introduce the features and key points of the “MAGNI” prototype, along with the elements that were the focus of our research.
Building a Living Open-World Environment
In “MAGNI,” we combined and utilized various features of Unreal Engine 5 to recreate vast natural landscapes with photorealistic quality.
In particular, we are striving to achieve both workflow optimization and expressive power, with the goal of “building open worlds even with small development teams.”
- Utilizing World Partition for automatic world segmentation and a level streaming system to efficiently develop open-world environments.
- Optimizing distant rendering with HLOD (Hierarchical Level of Detail) enables the rendering of vast landscapes.
- By leveraging Nanite, we can create high-quality, dense landscapes while reducing development time.
- Using dynamic lighting systems such as Lumen and Virtual Shadow Map (VSM), we create immersive worlds with time-based and weather-based changes.
- Leverage the Data Layer to achieve seamless stage transitions that maintain immersion.

![]() |
![]() |
![]() |
![]() |
To further improve development efficiency, we incorporated procedural generation tools such as Houdini and PCG (Procedural Content Generation).
Houdini was primarily used to generate large-scale terrain assets such as cliffs and caves.
For cliff details, we used mesh-based geometry. By extracting areas with slopes exceeding a certain threshold from the heightmap and converting them to mesh, we were able to add complex undulations to the terrain that couldn’t be achieved with the heightmap alone.
By utilizing procedural generation, it is possible to regenerate meshes that match new terrain without significant rework, even when heightmaps are modified due to level design requirements.
Regenerating large-scale assets like cliff meshes required significant time and multiple steps. To further streamline the workflow, we implemented a system using Jenkins nightly jobs to automatically update the meshes.
![]() |
![]() |
![]() |

The caves on the island are also generated using Houdini’s procedural modeling.
Using spline-based input, we could output everything from simple shapes to complex structures featuring stalactites and glowing minerals, enabling iterative adjustments between level design and final output.
![]() |
![]() |
![]() |

Unreal Engine 5’s new PCG feature was used for general-purpose asset placement to add environmental detail.
While the tool is simple, its editor-integrated responsiveness allowed designers to rapidly preview results and build complex placement rules independently.
The island’s forests were placed via PCG, which selected tree species and ages based on climate, elevation, and forest depth.
Then, based on the location where trees grow, a forest floor suitable for the tree species is formed.
![]() |
![]() |
![]() |
![]() |

MAGNI also features separate boss battle stages, built using Unreal Engine 5 and Houdini.

These boss stages use the Data Layer feature in World Partition, allowing them to be loaded within the same level as the open-world map.
This enables stage transitions while preserving character states, allowing for seamless presentation that maintains immersion without interruption.

Additionally, the areas where you battle bosses are modeled using Houdini.
We created rough shapes in Maya, imported them into Houdini for high-mesh conversion, and added details such as ice formations.
When creating large-scale background assets with a small team, procedural modeling proved immensely helpful. It allowed for faster reflection of changes on overall quality compared to manual sculpting and offered greater flexibility for content modifications.
![]() |
![]() |
![]() |
![]() |
![]() |
By combining these technologies, we achieved a balance between expressive visual quality and development efficiency, creating a high-end open-world environment.
Lighting and Visual Expression Utilizing HDR
In MAGNI, we leveraged the wide color gamut and high brightness range characteristic of HDR (High Dynamic Range) to deliver natural yet powerful visuals that span from deep shadows to brilliant highlights.
- Dynamic light changes via Lumen (day and night, indoors and outdoors, etc.) combined with Exposure and Eye Adaptation, enabling intricate light rendering that leverages HDR’s high luminance range and wide color gamut.
- Development based on HDR standards and consistent visual representation achieved through LUT (Look-Up Table) adjustments in SDR environments
As a result, MAGNI delivers a photorealistic and immersive visual experience—from the deep shadows of caves to the dazzling sunlight that bathes the landscape.
![]() |
![]() |
An Action Game Where ‘Movement’ Itself Is Fun
The setting is a vast island inspired by the natural landscapes of Norway.
We designed diverse movement actions—such as galloping on horseback, scaling cliffs, and dynamic traversal using axes and lightning powers—so that movement itself becomes a core part of the game’s enjoyment.
The island features multiple climate zones—snowfields, forests, and high mountains—with time and weather changing dynamically in real time.
The ever-changing scenery as you move enriches the exploration experience.

![]() |
![]() |
![]() |
![]() |
Dynamic Combat Leveraging Movement Actions
These unique movement actions are also utilized in combat.
Against the massive artificial giant made of mud and rock, ‘Mokkurkalfi,’, can be effectively attacked by climbing its body and striking its weak points.
Also, player can employ tactics derived from movement actions, such as hooking an axe onto a wild animal to pull it closer and attack.
This fusion of movement and combat delivers an intense, seamless action experience.
![]() |
![]() |
![]() |
![]() |
Storytelling and Cinematic Expression
The story follows Magni, a surviving warrior of the Vikings, and his loyal steed Gullfaxi.
Their peaceful life on the island is shattered when a mysterious calamity strikes: wild beasts become violent, and artificial giants made of mud and stone — the Mókkulkalfi — begin their assault.
To restore peace to the island, Magni and Gullfaxi must face this threat head-on.
Cinematic production incorporated:
- Motion capture
- Markerless facial capture from video
- Automatic lip-sync animation generated from voice audio
These tools helped bring authentic emotion and tension to each scene, enhancing immersion throughout the narrative.

![]() |
![]() |
![]() |
![]() |
![]() |
![]() |
![]() |
Performance Optimization for a Comfortable Play Experience
We optimized performance to ensure players can enjoy the vast landscapes and high-end graphics smoothly and comfortably.
- By understanding the characteristics of Nanite meshes, we reduced processing load without compromising visual or gameplay quality.
- By utilizing Virtual Textures, we efficiently rendered high-resolution textures and complex materials.
Through these optimizations, we achieved both high-quality visuals and stable performance.
![]() |
![]() |
Looking Ahead to Future R&D
In MAGNI, we were able to demonstrate the effectiveness of procedural generation tools like Houdini, in addition to verifying the capabilities of Unreal Engine 5.
We will continue our R&D of Unreal Engine and procedural generation technology, aiming to achieve both expressive power and development efficiency.
Eighting will continue to pursue the creation of more exciting and higher-quality games, with “challenge” as our guiding principle.
Gameplay Video
Event Exhibition
We exhibited a playable demo at the Sapporo Game Camp booth during Tokyo Game Show 2025. You can view highlights from the event here.

◆Sapporo Game CampOfficial website

Sapporo Game Camp is a joint event organized by Sapporo City and Sapporo-based game companies to foster IT talent and game creators in Sapporo and further invigorate the entertainment industry.
Our Sapporo studio is providing operational support.
※This research and development utilizes LiDAR data provided by the Norwegian Mapping Authority.
※The Unreal Engine logo is a trademark/registered trademark of Epic Games, Inc. in the United States and other countries.





































