Creating Cinematic Opening For Narrative-Driven Mystery Thriller Game In UE5

Introduction

Hi, I'm Francesco Furneri, an experienced 3D Texture and Lighting Artist based in Italy. I hold a bachelor's degree in Computer Science and a postgraduate qualification in Virtual Reality and Multimedia. I then moved to the UK, where I completed a Master of Science in 3D Computer Animation at the University of Kent (UK).

Over time, I've also collaborated on a range of commissioned projects for several studios, with an emphasis on crafting detailed 3D props for games and cinematic shorts.

In addition, I collaborated with TurboSquid and Shutterstock on articles and tutorials focused on 3D production with particular emphasis on teaching best practices in 3D modeling, texturing, and lighting through a series of lessons published on the Shutterstock Academy – you will find all the links on my LinkedIn profile, which will be included at the end of this interview. 

I have also worked as a 3D Lighting and Rendering instructor at a post-secondary institution in the U.S., and later served as an Unreal Engine mentor and consultant for a company in Italy. More details are available on my LinkedIn profile.

Last year, I had the pleasure of being featured in an interview on 80 Level, where I showcased part of a project I'm currently developing: a narrative-driven 3D mystery thriller with a strong focus on exploration and storytelling. The aim was to explain the overall workflow, from reference gathering to real-time rendering of the main location, along with the techniques used in the industry. In case you missed it, here is the link.

Thanks to 80 Level, I also had the opportunity to gain significant visibility and reach a wider audience through their platform by presenting additional interviews in the last few years.

For instance, if you're interested in procedural 3D material authoring, I've put together a detailed breakdown of how I created a fully procedural 3D steak in Substance 3D Designer.

Finally, my first contribution on 80 Level covered the complete workflow behind the creation of a detailed crocodile in Maya, ZBrush, and Substance 3D Painter. The article walks through each step, from visual concept and reference gathering to texturing and final rendering.

Redwood Game – State of the Art

Since my latest interview, the game project has evolved a lot, despite still being in the early/mid-stage of development. As mentioned earlier, the adventure is driven by mystery, as well as a strong narrative and storytelling. To briefly recap, the story goes as follows:

In my previous interview, I explored the mansion part and focused on what inspired me from the very beginning. I also discussed a few technical aspects, such as reference gathering, blockout, and working with a modular approach for floors, walls, and so on. Practically, I explained more about the inside of the villa, but the project has been enriched with additional locations and features.

In this article, I would really like to present the evolution of the game, covering new environments, cinematics, shots, and rendering, along with the main character's presence, as well as the user interface and starting menu.

The image above depicts the current key art of the mansion's exterior, setting the tone of the game with a subtle sense of mystery and an invitation to explore. I deliberately chose a low-angle shot in order to emphasize the majesty of the mansion, while framing the scene with two out-of-focus plants positioned opposite each other.

The color scheme follows a subtle orange and teal contrast to enhance visual depth and separation. Subtle greenish and bluish tones were introduced in the shadow areas to increase color contrast and create a more mysterious atmosphere – the cloudy sky also helps add a subtle, unsettling tone.

Finally, the viewer's eye is drawn towards the warm light through the window, establishing a clear focal point and reinforcing the narrative focus of the scene.

At the moment, the screenplay and most of the environment are complete, apart from specific game assets that need to be developed for key narrative moments, as well as additional locations that emerged during the writing process.

Besides, I'm currently working on the gameplay and programming side of the project, using both Blueprint and C++ depending on the requirements. This step involves implementing core gameplay systems, designing exploration and puzzle mechanics, and refining the overall narrative flow.

The Game Concept & Its Visual Direction

Apart from the plot explained earlier, this narrative-driven mystery thriller is set in an isolated villa, where the player explores the environment, from inside to outside, through puzzles and environmental storytelling. Essentially, the core of the game is based on exploration, interaction, inventory management system, along with other details that can't be revealed at the moment. Objects, lighting, and composition play a role in guiding the player and suggesting pieces of the narrative.

In addition, pivotal items can be examined at 360 degrees to find clues, and combined sometimes. For instance, the following clip shows a wooden piece being picked up and inspected during the gameplay – of course, a few game mechanics will be refined or slightly changed when the game is out.

This first-person experience is built around immersion and the player’s interaction with the environment. Each room, and more generally every space, has been designed to connect the player with the story: nothing is placed at random, and every detail plays a role in the narrative, while also enhancing the visual impact.

Moreover, the mood tells the story through lighting, color, and atmosphere. From a technical perspective, I used the decals in Unreal Engine for the storytelling of each location. Elements like dust, scattered objects, crumbling walls, and spider web – to name a few – make the difference. For instance, here is the basement before and after the application of decals: notice how it conveys a much stronger sense of atmosphere and realism with decals enabled.

Surface materials are crafted with a high level of detail to better reflect the environment the player interacts with.

From the beginning, the stylistic direction has always been that of a thriller rather than a traditional mystery (in the style of Agatha Christie) or horror experience, even though it’s still enriched with moments of tension. At the same time, the player is engaged in solving puzzles and unlocking new areas of the mansion. To choose the right direction, I've been developing a good amount of documents and media, including screenplay drafts, concepts, storyboards, flow charts, and game mechanics, all consolidated into a structured game design document (GDD) to support production – they are all essential for defining the core direction of the project.

The reference material was vital throughout the pre-production and helped speed up the workflow.

In particular, the blockout was crucial before building the final environment, as it served as an essential step to evaluate volumes and spatial relationships.

Early Development Features

Once I had completed the screenplay and the pre-production steps, I decided to move forward with the first cinematic sequence of the game because:

• It is the first sequence the player experiences, setting the initial tone;
• It establishes how the player reaches the mansion;
• It provides the opportunity to define the visual language and overall atmosphere of the experience.

The intro sequence follows the main character driving through the fictional Redwood City. It's a quiet day, with a few birds flying in the sky. A large sign introduces the player to the adventure, and the camera keeps the focus on the car as it approaches the mansion.

As the car comes to a stop in the backyard, we finally take a closer look at the character for the very first time: he steps out, looks around, takes in his surroundings –  this is where the adventure begins, and the player takes control.
 
Finally, the sequence is also enriched with a custom soundtrack and sound effects to reinforce the emotional impact and immersion.

There were essentially three locations I had to develop: the hill area and its surroundings, just for the cinematic, the outside of the mansion, and the backyard, and, of course, the interior of the mansion, which represents the core of the gameplay.

The hill was not sculpted inside of Unreal, apart from a few subsequent adjustments. It all started with World Creator, which is a software for building terrains and complex landscapes.

From World Creator, I exported a high-resolution heightmap of the hill, along with supporting textures like slope and flow maps – these are crucial while texturing the landscape in Unreal as they offer more control and macro variations in a few areas.

Inside of Unreal, after importing the heightmap of the hill and the supporting maps, I created a custom Landscape Material based on three repeated textures: grass, soil, and a ground variation with leaves.

The main challenge was to avoid visible tiling and repetition across large surfaces. This was addressed by introducing macro normal and color variations, blending masks and materials, and subtle texture offsets combined with texture bombing to break up uniform patterns and achieve a more natural result.

The set dressing also included the use of the Foliage Tool in Unreal, which allows scattering plants, trees, rocks, and stumps throughout the environment with ease.

In addition, the presence of external props and distant meshes, like houses and mountains, helped complete the scene, filling empty areas and extending the perceived depth of the environment without compromising performance.

The two compositions show the presence of natural elements and props with a convincing scattering and distribution.

A simple word of advice: take time to observe real-world environments, and notice how randomness exists in nature. If your landscape looks artificial or too industrial, it's time to make some changes!

The most important aspects of the development were:

• A modular approach for the main building, where repetitive elements such as roofs, walls, and floors are built from modules. To break the visible tiling caused by repeating the same elements, several techniques were applied. I invite you to check the Modular Approach section in the Mansion Interview for a closer look;
• Procedural Ivy Generation applied across walls, terraces, corners, and so on. I generally use SpeedTree in my projects, but for this case, I switched to Dash, which has become very popular in game environment production. This awesome Unreal Engine plug-in allows you to draw ivy directly along surfaces, achieving a natural look through a set of customizable parameters.

• Material Blending for specific terrains and surfaces. Some materials were built and blended directly in Substance Designer, while others were blended in Unreal Engine by using vertex painting techniques and a Dash feature;
• Decals were used extensively throughout the game, particularly on walls to add weathering and damage details.

The Sequencer & Its Useful Features

Let's get straight to the intro. During the production of this cinematic scene, I leveraged the Unreal Engine's Sequencer to achieve the desired result.

The sequencer allows you to add and compose several shots in your level like a real director. It is also a flexible and time-saving tool across the shots. Each shot has its own set of elements/components that can be added with ease.

For my purposes, I had to consider different components, such as:

• Actors representing any dynamic or interactive elements in the scene. For instance, the main character, the car, and the bird system blueprint fall into this category;
• Camera for composition and shot framing;
• Lighting to reinforce mood and visual clarity. For some shots around the villa, I had to include specific light sources to drive the viewer's eye;
• VFX was produced in Niagara to simulate dust on the ground, water splashes near the car tire, and other similar effects;
• Subsequences that represent other sequences embedded within a shot. The car animation is an example of subsequence, and the way I animated it will be explained in a few seconds – it turned out to be a major time saver during production.

You can add as many elements as you like within your shot, and the real flexibility comes from the ability to keyframe a wide range of properties over time.

Let's take a closer look at one shot, made of 113 frames, to comprehend how it works. The camera is low and the car moves, kicking up dust from the ground.

The highlighted items represent some of the elements added to the shot. The horizontal bars indicate the time span during which each element is active in the timeline.
Finally, the properties listed under each item are the ones that can be keyframed: some of them come with the item, like Transforms, while others can be added by pressing the "plus" button beside the item itself.

To be honest, I had to experiment with different properties to determine which were critical to achieving the final result. Consequently, I would like to mention a few of them.

• Transforms. This property allows changing the actor's location, rotation, and scale over time by keyframing its transforms. It was also useful for freezing the actor's transforms at the beginning of the shot. For instance, the license plate was fully customized and imported as a separate mesh – its initial transform was keyframed based on the car's resting position.
• Camera Shake. A very useful camera property that relies on a custom Blueprint class, derived from the CameraShakeBase parent class. This custom class can then be applied over a specific time range within the timeline (BP_CameraStartEngine). The effect is self-explanatory as it creates a shake, defined within the Blueprint Class.

• Attach. I used this property for various purposes, primarily to link actors together and maintain relative transforms during animation. The Attach property establishes a parent-child relationship: For the license plate, it defines the car it is attached to, keeping the plate bound through the animation. For the car, I created an empty Actor called Translator2 and attached the car to it. The goal was to control the car's height, as the original baked animation caused the wheels to partially intersect with the ground, and I couldn't easily modify the keyframes – the Translator2 actor came in handy! 

• Events. That's a very powerful concept while working with the Sequencer, as it allows you to trigger custom events at a specific location in the timeline. I used it extensively in the cinematic for different purposes: Spawning VFX like dust at a specific frame and spawning the bird flock. Without using Events, it wouldn't be possible to control when a specific action occurs during the sequence – by simply dropping an actor in the shot, its animation would occur immediately, but that’s not always what we are looking for.

The diamond-shaped marker can be moved in the timeline and is associated with a custom event in the Sequencer Event Graph, where we can define any logic and behavior.

Subsequences

To speed up the development of the game intro, I decided not to animate the car frame by frame. Instead, I relied on a powerful tool in Unreal Engine that allows you to record actions during gameplay and bake them as subsequences in Sequencer.

This is called Take Recorder under Window/Cinematics. I primarily used this feature for car animation.

Once it is selected, you can specify the source, which is the Blueprint of the car (BP_Sedan_Chaos) – as a side note, the car was part of a free Vehicle pack volume from Switchboard Studios and was customized in terms of materials and textures.

The amazing thing about this package is that each vehicle can be driven in-game, as it is built on Unreal Engine's Chaos Vehicle system.

The car is fully rigged and comes with built-in suspensions, which greatly enhance realism during the animation.

This setup allows the car to naturally react to the terrain, adding subtle secondary movements that would be time-consuming to animate manually. As a result, it significantly streamlines the animation process while maintaining a high level of physical credibility.

Getting back to the Take Recorder feature, which allowed me to pilot the car throughout the shots and use the recorded track as a subsequence.

Let's see more about that.

Once the car is positioned in-game, you can simply press the red record button to start recording, as if you were playing a car racing game. A countdown then notifies you before the recording begins.

At this time, Unreal records what is happening during the game and saves the key frames as a new subsequence – by pressing the ESC button, you stop the recording.

This is a wonderful time saver for making cinematic scenes!

When you are done, remember to import the new video sequence in the timeline and play the subsequence as shown in the previous video. The myriad of red dots below represents the baked keys, and the small frame on the right shows the capture from the player’s point of view. That's awesome, isn’t it?

By using the Take recorder feature, you don't need to animate the car using traditional key frame animations – the only requirement is a solid car rig compatible with Chaos Vehicles and a bit of patience while recording a gameplay session.

Cinematic Direction

Using the tools available in Unreal Engine’s Sequencer is only half the work. To create a game intro that truly stands out, a strong foundation in direction makes all the difference.

Personally, I've always had some experience with shot composition and camera work, so I used this knowledge to orchestrate the cinematic.

These are some of the main camera movements, some of which I used in the game intro sequence.

Let's take a closer look at some of the directorial choices behind the sequence.

In the first three shots, for instance, the camera moves slowly, introducing Redwood to the player.

The opening shot sets the tone and immerses the player in the environment, with a very slow camera push-in that reinforces the viewer's connection to the scene – technically referred to as a Dolly-in.

Then, part of the environment is revealed through a Truck movement to the right – second shot. This motion enhances immersion by drawing attention to the plant animation, further emphasized by the use of depth of field.

The third shot begins with a Pedestal movement from top to bottom, followed by the car entering the frame from the bottom-left corner.

When observing the three shots as a whole, a subtle and cohesive camera progression becomes evident, guiding the viewer from an elevated perspective down to ground level: the intention is to give the player a thorough overview of the scene first, and then gradually shift focus toward the ground, where all the action takes place.

The shot where the car pulls away and continues along the road is characterized by a low-angle shot. This perspective enhances the subject’s sense of scale and presence, making it appear more powerful. The use of camera shake, motion blur, depth of field, and dust effects further amplifies the sense of motion.

Jumping to the shot where the car drives through the small rural village, the camera performs a Truck movement to the right. However, an interesting aspect to consider is the clear separation of spatial layers into foreground, midground, and background.

The foreground is described by nearby buildings and flocks of birds, enhanced by motion blur and supported by off-screen ambient sounds, which contribute to a stronger sense of immersion. The midground features the car passing through the scene without drawing excessive attention to itself, maintaining a balanced visual hierarchy. Finally, the background is defined by the distant hills and surrounding nature, providing depth and contextual grounding to the shot.

A word of advice: a common mistake in this type of shot is matching the camera movement too closely with the car. If the camera follows the vehicle in perfect sync, the result can feel staged, almost like a chase, which takes away from the natural flow of the scene

Instead, the camera should maintain its own motion, allowing the car to move through the frame rather than being constantly tracked. This helps preserve a more organic feel and keeps the focus on the environment as a whole, rather than isolating the car as the only point of interest.

Focusing more on the shots of the villa, the first three or four sequences are essential in introducing the player to this new location. The opening shot uses a low camera angle combined with a longer focal length, giving the mansion a majestic and imposing presence while preserving a narrow field of view.

The depth of field is focused on the gate, drawing attention to the ivy and the subtle particles floating in the air. At the same time, the camera performs a slow pedestal movement upward, gradually revealing the villa and reinforcing its imposing appearance.

The next three shots are similar and designed to reveal new details of the mansion. 

Particularly, in the second shot, a similar camera setup is maintained, but combined with a left-to-right truck movement. The presence of the statue introduces a sense of mystery, revealing only a portion of the mansion instead of exposing it all at once. The use of depth of field is reversed: this time, the ivy falls out of focus, shifting the viewer's attention towards the side of the building.

Almost at the end of the intro, another interesting camera movement is introduced.
In a couple of sequences, the purpose is to guide the player's eye towards specific actions due to their narrative importance within the cinematic.

As the scene reaches the backyard, the intention is to have the player follow the car's final movements before it comes to a complete stop. This is achieved through a controlled camera pan.

The movement consists of a rotation around the camera’s vertical axis (Z axis, assuming Z as the up vector), allowing the camera to track the car for roughly 30 to 45 degrees.

It's also worth noting that, during this sequence, the main character briefly appears through the car window: his face is never fully revealed, maintaining a sense of ambiguity that has been carefully preserved since the beginning of the cinematic.

When the main character finally steps out of the car and looks around, he takes a few steps forward with the camera following his movement: a gentle camera pan keeps the viewer's attention anchored to him.

Bringing the Main Character to Life in Redwood

A game works best when its characters feel believable and convincing. The main character plays a crucial role in this, guiding the narrative while establishing an emotional connection with the player. He's the nephew who received the villa as an inheritance from his uncle.

Despite being a first-person game, where the character is not constantly visible, his presence is very fundamental to the narrative: the player should build a connection with him, understand his motivations, and feel immersed throughout the gameplay and real-time cinematic sequences.

The idea behind my character is quite simple: I've always liked having a protagonist who fits the role – young but smart, someone capable of dealing with situations instead of feeling passive or inexperienced, and this is what I obtained.

MetaHuman Creator is an amazing tool for producing high-quality characters for games.
It's helping me a lot during the game development as it speeds up the whole creation process.

At the time I’m writing the article, I'm still using Unreal Engine 5.4, and MetaHuman is just available as an external tool via web browser: I will soon upgrade to version 5.6 to take advantage of the built-in MetaHuman system integrated directly into Unreal. Working with MetaHuman Creator is pretty straightforward: the customization level is important, and it helps speed up the process without starting from scratch.

The avatar creation process starts from a generic MetaHuman model and allows changing several face, hair, and body features.

For instance, you can customize your hairstyle, choose whether to add a beard, define body proportions, and even use the sculpt mode by manipulating simple control points. Every action affects the viewport in real-time, providing immediate feedback. That's pretty cool, isn't it?

Furthermore, you can choose the outfit from a preset of MetaHuman clothing. If you prefer not to spend much time customizing it, the default options provide a solid starting point. On the other hand, if you prefer to have more control over your clothing, you can always think about that at a later stage, and apply it to your character afterward – tools like Marvelous Designer are perfect for creating tailored garments.

In my case, I found that sweater to be the perfect garment for my character: just a couple of customizations like colors and patterns inside the editor, and that's it!

I believe this style is exactly what I was looking for: simple, effective, and with a subtle '90s vibe. In fact, the game is roughly set between the '90s and the early 2000s, although it's not explicitly stated. Once the character is finished, we can jump into Unreal Engine and import it into the project. The great thing is that the character comes with a standard Control Rig ready to be used for animation.

Here you can make a decision depending on the type of animation you're creating.

You can use the Control Rig as it is to animate the controllers and work in a sort of Graph Editor, much like in Maya. Keyframe-based animation is inherently time-consuming, especially for complex sequences, but it works for shorter ones when precise control is required.

You can opt for external animation libraries such as Mixamo to quickly generate and apply animations.

Generally, the skeleton from Mixamo is not ready to use out of the box because it requires matching the Metahuman skeleton somehow. Fortunately, Unreal comes in handy with the IK Rig and IK Retargeter tools, which match the bones between the source and target skeletons.

Another professional solution that I recommend is Rokoko: a motion capture system designed to produce realistic animation data, which can be streamed or exported and then refined inside Unreal Engine through retargeting and Control Rig workflows. Beyond motion capture tools, Rokoko offers both free and paid mocap assets that can be integrated into your workflow – even for commercial purposes.

In the Redwood cinematic intro, I used both keyframe-based animations and some free mocaps from Rokoko, cleaned up and retargeted for MetaHuman. In the sequence where the main character watches the crow fly away, I used a mocap clip representing a skeleton walking forward.

To make it possible, you can add any Animation Clip to the Sequencer and set the exact moment where you want the animation to start: working animation clips in the sequencer is awesome, as it allows you to blend more clips together.

Moreover, once the animation is imported, you can refine it using Control Rig by adjusting controllers, polishing poses, and refining timing directly in the Sequencer. You practically start from a good animation and adjust it as you like in different ways – additive animation layers are also a good choice in this case.

In Redwood, the mocap animations I used are already quite solid and only require minor polish, such as improving foot placement and resolving hand intersections; there are no additional layers involved at the moment.

A simple trick I used in these sequences is reusing mocap data across multiple shots. Some mocap animations can be quite long and varied – such as the walk cycle in my case. For this reason, I used the Sequencer to isolate specific frame ranges and reuse them across different shots. Keyframe animation was used in certain cases, such as in the shot where the character shifts into gear and drives off towards the villa.

When you don't find a mocap animation that suits your scene or when the action is short, you can always decide to use the Control Rig to create manual keyframes: in this sequence, the character simply lowers his hand and shifts into gear, with the torso rotating to the right. The curve Editor gives you full control over your animation by working with curves.

What about face animation? Like body animation, the face rig comes with MetaHuman and offers a very powerful set of options. At first glance, the interface might appear a bit confusing due to its myriad of controllers (the yellow dots).

The face rig is quite simple, but it takes time to create a complete animation. All the facial expressions you see in the game intro are created by keyframing those yellow dots. The video below shows an example of how the facial rig is solid and robust.

Nonetheless, I'm going to adopt another solution for facial animations in the future because the real-time cinematics will become more prominent and demanding. Within Unreal Engine 5.6, for instance, users can capture facial expressions, eye and lip movements, directly from a common webcam.

On the official Unreal Engine website page, the community section features a tutorial on setting up webcam-based facial animation for MetaHumans. I invite you to take a look at the video for more details.

Once the recording is complete, we can bake the animation and use it in the Sequencer: the process is the same as before.

To sum up, this workflow is well-suited for Redwood, as the game features several real-time sequences, and strong character performances are critical for a narrative-driven experience. In the next section, we'll wrap up the cinematic aspects and take a look at some rendering settings that help create clean and professional-looking cinematic renders.

Rendering the Intro Cinematic

While creating the final video, you need to decide whether to use real-time rendering or a higher-quality offline approach. As you know, since UE version 5, the game engine uses Lumen as a global illumination system that is based on the Ray tracing algorithm – more info here.

The default setup relies on Software Ray tracing, but you can switch to Hardware Ray tracing for better results if your video card supports it. However, Unreal also offers another method for rendering sequences: Path Tracing. It is typically used for offline rendering (e.g., Arnold) as it requires significantly more computational resources, but it delivers higher visual fidelity, especially for film.

Nonetheless, real-time ray tracing should not be underestimated: despite being a biased approach, it can still deliver excellent visual quality for cinematics when properly configured.

In Redwood, I chose Ray tracing for the intro cinematic, as it offered a solid compromise between high visual fidelity and practical rendering times, which was essential during production. For the gameplay, Redwood leverages Lumen's software ray tracing to support fully dynamic lighting, which fits well with a narrative-driven experience.

The player can freely switch lights on and off, while the global illumination system automatically updates the scene, preserving consistency and realism.

The cinematic intro required a different rendering setup compared to the gameplay counterpart.

For the intro, the video is composited, exported, and played as .mp4 during the game, before the player takes control over the main character. The quality had to be as high as possible, even at the cost of longer rendering times.

For the gameplay, the focus shifts to real-time performance. Concepts like texture streaming, LODs, HISM, and appropriate anti-aliasing methods make sense in this case.

Finally, there are also other common settings involving Lumen and the post-processing aspects in Unreal Engine to consider.

Let's take a closer look at a few settings for the intro cinematic. The Movie Render Queue is the right solution for rendering sequences in Unreal Engine.

The window allows you to add specific options – via the plus Setting button – which influence the rendering quality. Depending on the shot being rendered, I created a few presets with these specific settings enabled: Anti-aliasing, Console Variables, and Game Overrides, used in a couple of shots.

I decided to use a .png sequence as output instead of an .exr one, because I had previously managed most of the color grading and post-processing aspects within the Unreal editor. A word of advice: if your intention is to have a thorough control over the look of your cinematic in post-production, consider exporting it as an .exr sequence.

The Anti-aliasing section is crucial, as it includes two fundamental parameters to set up:
Spatial Sample Count and Temporal Sample Count.

The former increases the number of samples per pixel evaluated per frame – much like a traditional supersampling approach – improving image quality within a single render. The latter accumulates multiple sub-frame samples within the same frame, helping reduce noise and improve image stability, especially in motion blur, camera movement, and flickering scenarios.

In my shots, since there are situations with motion blur or depth of field, the temporal sample count had a relevant impact there: for instance, the shot featuring the ivy on the gate in the foreground required more temporal samples – raised from 2 to 8 – to reduce the flickering caused by camera movement.

A word of advice: spatial and temporal sampling already produce clean and sharp results, so additional anti-aliasing methods are often not necessary. In fact, I set the Anti-Aliasing Method to none, which proved to work best for my cinematic intro. In fact, techniques like TAA and TSR can slightly blur the image and soften fine details. However, the anti-aliasing method is still useful in real-time or when using low sample counts, as it helps reduce artifacts and stabilize the image.

To tweak the quality of some rendered shots, I also introduced a few Console Variables, though not always necessary.

You can modify scalability groups, and most importantly, disable texture streaming: this ensures that textures are fully loaded into memory, helping to avoid visible popping during rendering. If you don't need this level of customization, you can alternatively add the Game Overrides setting, which automatically sets a Cinematic Quality setting and disables Texture Streaming by default.

The game uses the Post Process Volume to enhance the overall look of the shots. As stated earlier, I didn’t need to tweak the scene in post production so much – apart from small level adjustments and curves in Davinci Resolve – as I already obtained what I was looking for, directly here in the editor.

When working with the Post Process Volume, you have a bunch of options to consider. For the intro cinematic around the villa, I tweaked a few parameters.

Bloom was used to create a soft glow, allowing bright areas to bleed into surrounding pixels; I set it to 1.6, keeping the effect subtle but still noticeable. Exposure played a key role in balancing the overall brightness and ensuring that highlights and shadows were properly defined. Color Grading was applied to enhance the mood and set the overall tone of the scene.

Moreover, color temperature was adjusted to achieve a warmer tone, while a subtle greenish-blue tint was introduced in the shadows to create a more cinematic contrast. The Film section, which is often overlooked, allows you to fine-tune the final image, helping to shape contrast, highlights, and overall visual response. The most important parameters are:

• Toe: controls how shadows are lifted or compressed. I used this option to make shadows a bit brighter;
• Slope: defines the overall contrast of the image;
• Shoulder: controls how highlights are compressed and rolled off, preventing unnatural clipping in bright areas.

The Main Menu & UI: Establishing the First Impression

In this last part of the interview, I would like to briefly discuss the main menu and the UI that the player can interact with from the very beginning.

At the time of writing, it is still a work in progress, but it already gives a clear idea of the direction I want to pursue. What I really enjoyed while designing the game menu was the possibility of integrating the user interface into a dynamic, real-time scene, with the camera gently shifting from left to right in a loop.

The scene is based on a room from the game environment and features an old desk with various objects.

As a brief side note, apart from a few props that I created myself, some of the assets used in the scene were sourced from Sierra Division and purchased with a proper license. Sierra Division is a professional 3D asset studio known for high-quality production-ready content.

Returning to the main menu, the structure is quite simple as it should be. The Story section allows the player to start a new game or load a previous save, the Options section includes a list of customizable game settings, and the Credits section includes information about the people, tools, and resources involved in the project. Finally, the Quit option allows the player to exit the game.

Here is an example of Video Settings where some aspects can be set up depending on the player’s hardware. This type of menu provides flexibility and control over the experience.

All of these settings are currently saved and restored when the player plays the game.
In the image below, post-processing effects are disabled on the left and enabled on the right.

This allows players to tailor graphical scalability to their hardware and preferences.

What happens when the player starts the game?

First of all, we are rendering a diegetic/cinematic main menu which includes the scene in real-time. Many modern games use dynamic elements to introduce the player to the experience, and this approach works particularly well for a narrative-driven experience.

When the player clicks on Story, they can start a new game or restart a save: at the moment, there is no saving functionality yet, but it will be implemented.

By clicking on New Story, you are asked to proceed: this is the moment where the cinematic part takes control, and the camera starts moving towards the desk.

Here is the first part of the camera animation in action.

The camera slowly moves towards the table, and a few dynamic elements respond. The candle goes out, producing wispy smoke for a while, and the book on the right opens, revealing the first two pages.

At this point, the player must click on Read Book to proceed with other interactions. The book opens with a brief introduction, while giving the player the ability to interact with its pages, read transcriptions, and explore its contents at their own pace.

The menu, as well as the events associated with the interactions, have been developed in C++ by creating appropriate class structures. For instance, I developed a few C++ functions and classes for maintaining the page count and updating the widgets on screen.

Transcriptions are essential for revealing the content of each page. They appear in almost every note throughout the gameplay and are designed to make the text clearly readable.

I managed a C++ class by defining an array of transcriptions – one per page – retrieved from a String Table structure in Unreal Engine. This structure allows you to store key–string pairs that can be efficiently accessed from C++ code.

In fact, this macro gets the content that corresponds to the inKey, and makes it available in C++ code.

#define GET_DESCRIPTION(InKey) LOCTABLE("/Game/UI/StringTables/ST_MenuBookPageContents.ST_MenuBookPageContents", InKey)

For example, Page6 is used as a key to access the corresponding transcription content.

Conclusions

This brings us to the conclusion of this extensive interview, where I had the possibility to discuss the current state of the art, together with several technical aspects.

At the moment, the presence of a game menu and a functional user interface already provides a clear foundation for shaping the player experience, even though the project is still evolving. The cinematic intro plays a vital role as it helps establish the tone, guiding the player into the narrative and setting the emotional context of the experience.

Furthermore, to conclude the cinematic intro, it took me around two months, covering environment setup, lighting, shot composition, and rendering: the hill and its surroundings had been previously modeled for another personal project, but it was never showcased before.

One of the biggest challenges is that I am developing the game on my own. This means handling different aspects of the project, from environment art and lighting to storytelling and technical work. It also requires staying organized and making decisions without immediate feedback. At the same time, working alone gives me full creative control, allowing me to shape the experience in a consistent and personal way.

From a technical point of view, the main challenge was establishing the right mood of the game in just 2 minutes of intro, which meant carefully balancing tones, lighting, and composition.

A further challenge was related to the overall screenplay, which required careful structuring to ensure clarity and engagement in the player's actions: writing a solid and intriguing story is never predictable; it involves a lot of trial and error, constantly building and reworking the narrative.

Being a 3D artist with several years of experience in the industry across the full pipeline, spanning both art and programming, I feel confident that I can achieve what I have in mind for this project.

For people who desire to develop their own game, my advice is to proceed one step at a time, focusing on small, achievable goals and gradually building up the project… There is always time to iterate and make things better as the project evolves.

Now, the question is "What to expect from the future?" As I mentioned, the environment is almost done, apart from a few new rooms to be considered, and the screenplay is strong and well-detailed. I'm currently developing the programming in Blueprint and C++ and implementing the game mechanics and the story progression:  I know it's a lot of work, but I’m fully committed to it, and I'm sure you are going to like it, even though I can't reveal more details at this stage. To sum up, the game is currently being developed and is planned for release on Steam in the coming years.

I also invite you to follow me on my ArtStation and LinkedIn profiles, where I'll be sharing updates, teasers, trailers, and further insights into the project.

Francesco Furneri, Senior Texture & Lighting Artist

Interview conducted by Gloria Levine