How to Design Emotional Game Environments for Sky: Children of the Light

3D Artist Flora Yu, currently working at thatgamecompany on Sky: Children of the Light, discussed her approach to real-time environment art, emotional worldbuilding, and performance-conscious level production. In this interview, she breaks down her work on Season of Two Embers and Season of Duets, explaining how wayfinding, spatial composition, and optimization planning shape player experience in a large-scale live-service game.

Introduction

Hello, my name is Flora Yu. I'm a 3D Artist specializing in real-time environment art and worldbuilding, currently working at thatgamecompany on Sky: Children of the Light. My work on Sky focuses on environment art for a large-scale live-service game, including building stylized spaces and assets, translating art direction into real-time levels, refining lighting, materials, composition, and readability, and collaborating with designers and engineers to make sure the final experience works across platforms.

Because Sky is a live game, my work is not only about creating beautiful environments, but also about making them functional, performant, and emotionally clear for players. Before that, I worked across games, animation, and commercial visual production. I contributed to RoboCo, a VR sandbox game focused on engineering and STEM learning, where clarity, interaction, and performance were central to the production.

I also worked on Ink: Mountains and Mystery, a game inspired by traditional Chinese landscape painting and developed in collaboration with the Palace Museum, where my role focused on character and environment-element rigging, as well as animation for characters and scene props. In addition, I worked with Chamonix Vision on stylized commercial animation projects, exploring cinematic lighting, rendering, and art-directed 3D production.

These experiences gave me a broad technical and artistic foundation before I began focusing more deeply on game environments. My path into games started from 3D animation. I studied 3D animation and digital art at the School of Visual Arts in New York, and at first I thought I would enter the animation industry. During my junior year, I took a game design course, started learning Unity more systematically, and also gained internship experience in the game industry. That changed the direction of my career.

I became interested in real-time production because games allow artists to build worlds that players can actively move through, interact with, and remember through their own experience. Later, I pursued a master's degree in Learning Design and Technology at Harvard because I wanted to better understand how people learn, engage, and find meaning through designed experiences. That background still influences how I think about games and game spaces today.

I see games as an artistic medium with the added power of participation, and environment art is where I feel I can bring together visual design, technical execution, and meaningful player experience most directly. That is the direction I continue to grow in today: creating game environments where visual design, cross-disciplinary problem-solving, and meaningful player experience come together.

Environment as Wayfinding and Storytelling

Environment design is one of the most powerful ways to communicate with players without speaking to them directly. A strong game environment has to do more than look beautiful. It has to work as a playable space, carry the narrative identity of the world, and guide the player's emotional response at the same time. Players need to understand where they are, where they can go, and what matters in the space, but that guidance should feel natural rather than overly instructional.

When I design or build an environment, I usually start with the player's first read. In the first few seconds, what does the player understand about the space? Where does their eye go first? What feels safe, mysterious, dangerous, or inviting? A lot of that comes from visual hierarchy. I often look at a scene almost in grayscale before relying on color. The strongest value contrast, the cleanest silhouette, the brightest light, or the most open composition will naturally draw attention. These are useful tools for directing movement because they allow the environment to guide the player before any explicit instruction is needed.

For wayfinding, I think in multiple scales. At the macro level, players need strong landmarks, silhouettes, or spatial anchors that help them orient themselves from a distance. At the mid level, they need readable paths, material changes, elevation changes, framing, and repeated visual motifs that explain how one area connects to another. At the micro level, smaller identifiers such as color grouping, prop dressing, signage, lighting pockets, or unique shapes help players recognize specific places. When these layers work together, players can gradually build a mental map of the environment.

To me, that is the ideal kind of guidance: the player feels like they are figuring things out on their own, but the space has been quietly supporting that discovery all along. That discovery can be spatial, like understanding a route through a city, or narrative, like noticing traces of who lived there and what happened before the player arrived. When the environment gives players enough visual cues without over-explaining itself, the experience feels more personal, immersive, and rewarding.

The emotional side of environment design is closely tied to storytelling. I often approach an environment almost like a character study. A character has a history, a personality, a role in the story, and a recognizable visual silhouette. A strong environment needs a similar kind of identity. Its function, history, spatial structure, dominant shapes, palette, lighting, materials, density, and atmosphere all contribute to how players read it. Once that identity is clear, the environment stops being just a location and starts to feel like a place with memory, purpose, and personality.

A strong example from Sky is the city area in Season of Two Embers. The city needed to feel layered, lived-in, and full of history, but it also had to stay clear enough for players to move through naturally. In denser areas like the market, the team had to solve a practical navigation challenge: many tents, stalls, and narrow paths shared similar shapes, so the space could easily become disorienting.

The solution was not just to add direct signs, but to build orientation into the environment through taller structures, recognizable landmarks, hanging decorations, overhead elements, and color grouping. These elements gave players visual references above and around the crowded tent level, helping the market feel active and layered without becoming confusing.

My own focus in the surrounding living areas was more closely tied to narrative atmosphere and the history of the city. Since the main street already gave players a clear route, the side roads and residential sections could support exploration without making players feel lost. I wanted those spaces to feel like they belonged to real inhabitants, with traces of daily life, local materials, small props, fabric movement, lighting, and environmental dressing that suggested how people lived there and what kind of past the city carried.

In that part of the work, navigation was still important, but the larger goal was to make the city feel emotionally grounded rather than just functional. When those layers work together, the environment can support movement, communicate story, and shape mood without making the design feel too obvious. Players can read the world through visual cues, but still feel that they are discovering it for themselves.

Building the Season of Duets Concert Hall

I usually think of spatial composition, pathing, scale, lighting, color rhythm, and dressing as one connected system. They are all tools for shaping an emotional space, but each one solves a different problem. Pathing shapes how players enter and move through the environment, composition controls how the space reads from different viewpoints, scale determines how the environment feels around the player's body, lighting and color define the emotional focus, and dressing gives the space texture, warmth, and a sense of life.

A good example is the concert hall I worked on for Season of Duets in Sky. The experience begins before players even see the hall. They arrive by boat through a narrow canal, surrounded by a more intimate, grotto-like environment. From the layout and pathing stage, that entrance was designed as a transition. The player is carried through a smaller, quieter space first, almost like a gentle water-ride sequence, which slows down the pace and builds anticipation. When the canal finally opens into the larger concert hall, the reveal feels more memorable because the player has moved from compression into release.

Once players enter the hall, the layout shifts from a guided arrival to a space of choice. I did not want it to feel like a single fixed viewing area. Instead, the seating and circulation were arranged across different zones and levels, so players could choose how they wanted to experience the performance. They could sit near the water, gather around slightly elevated stone tables, stay along the arched corridor, or discover higher and more hidden viewing spots. Those choices gave players small moments of agency inside the space, while still keeping the overall layout clear and centered around the concert.

Composition was built around the central stage. Because this is an interactive space, the composition could not depend on one perfect camera angle. Players might be sitting low near the river, standing in a corridor, moving along the side paths, or watching from above. From each of those positions, the stage still needed to remain legible as the heart of the room. The layout, sightlines, and surrounding architecture were arranged so players could clearly see the performance area, while also seeing enough of the audience and environment to feel the atmosphere of a shared concert.

Scale was one of the most important parts of making the hall feel comfortable. The space needed to hold a large number of players without feeling crowded, but it also could not feel cold or empty when only a few people were there. The larger hall provided enough presence, while the smaller seating areas, tables, candles, and nearby props brought the scale back down to something personal and approachable.

That sense of scale is especially important in Sky. Many objects in the world are designed closer to the scale of spirits or adult figures, which can make the child-like player avatars feel small. For this concert hall, I wanted the social details to feel more tailored to the Sky kids themselves. The seats, tables, candles, and nearby dressing elements needed to feel like they belonged to the players, not like oversized objects in a space built for someone else. That helped make the hall feel welcoming and suitable for players to spend time together.

Lighting and color rhythm carried the emotional arc of the space. The entrance starts quieter and more enclosed, then opens into a brighter, softer, more romantic atmosphere. The stage was given the strongest light hierarchy in the hall: the skylight above, the natural beam coming through it, the caustic effects, and the dreamlike performance lighting all helped make the stage the highest-contrast and most luminous point in the room. The surrounding areas were kept quieter by comparison, so the performance felt ceremonial and special without making the rest of the hall feel visually inactive.

That rhythm helped control attention and mood, from arrival, to reveal, to gathering. Dressing was the layer that helped bring warmth and human presence back into the space. The base environment was largely stone: grotto walls, stairways, corridors, tables, and seating, so without contrast, it could easily feel cold or hard. To balance that, I used softer materials and smaller player-scale details to break up the heaviness of the stone and make the hall feel prepared for music, gathering, and shared memory.

Fabric, tablecloths, and curtains softened the hard surfaces and added gentle movement; candlelight made the seating areas feel more intimate; grass and flowers helped blend the architecture back into the natural river environment; and small props suggested care and use. Together, these details transformed the concert hall from a stone interior into a place that felt romantic, inhabited, and welcoming.

For me, the concert hall was not just a room for watching music. It was designed as a sequence: arriving through water, experiencing the reveal, finding the stage, choosing a place to sit, and sharing that moment with other players. When pathing, composition, scale, lighting, color, and dressing all support the same emotional goal, the environment becomes part of the performance itself.

Balancing Visual Ambition and Performance

I think in a large live-service environment, poetic visual design is not something you add at the end. It has to be designed within performance budgets, visibility planning, asset management, the player camera and likely viewing angles from the very beginning. The technical side is not separate from the artistic side; it is part of how the final feeling of the space becomes possible in a real game.

This became especially clear when I worked on the city and living area for Season of Two Embers in Sky. It was one of the largest and most detail-heavy environments I had worked on in Sky at that time, and the goal was to make the city feel layered, lived-in, and full of history. I also took on a major part of the environment performance planning and optimization work for this season, so I had to think about visual ambition and technical cost together throughout the process.

For a level of that scale, optimization cannot be treated as a cleanup pass at the end. As soon as we have a general direction for the level, we need to start thinking about performance limits, especially on lower-end devices: how much can be visible at once, where players will spend most of their time, and which areas truly need the highest level of detail. The first major decision is always where the budget should go. In my work, I usually prioritize the main player path, key gameplay spaces, and narrative-focused areas, because those are the places that carry the player's experience most directly.

Optional exploration areas can be more controlled or revealed more gradually, while background architecture may only need to support silhouette, depth, and the overall composition. Not every asset needs collision, baked lighting, complex materials, or the same amount of dressing. The role of each asset should determine how much technical and visual weight it receives. That planning becomes especially important in a dense city environment, where the cost is about how much geometry, dressing, collision, and lighting information can be visible or active from the player's camera at the same time.

I used the city layout itself as part of the optimization strategy. Walls, street turns, terrain shapes, larger buildings, and tent structures helped create a more layered and believable city visually, but they also controlled sightlines and worked with occlusion so hidden geometry and dressing would not need to be rendered unnecessarily. In that sense, an architectural element was not only worldbuilding; it could also become part of the performance strategy.

The tents are a good example of how narrative dressing and optimization can support each other. I wanted tents and residential corners to contain smaller props that suggested the owner, function, or daily life of that space, because those details helped make the city feel inhabited. But those details should not be rendered all the time from every angle. The tent structure, occluders, and visibility planning allowed us to keep storytelling details where they mattered, while preventing them from becoming too expensive when they were hidden or far from the player. The goal was not to remove detail for optimization, but to make sure detail appeared where it had the most narrative and emotional value.

I also made different decisions based on what each asset needed to do. Objects close to the player, objects on the main path, or objects tied to important story moments needed more careful modeling, LOD setup, collision, baked lighting, and material treatment. Background buildings or distant architectural pieces could often be instanced, non-collidable, non-baked, or simplified, because their job was to create depth and atmosphere rather than direct gameplay. A strong environment is not built by treating every asset equally. It is built by understanding what each part contributes to the player's experience.

That is how I think about balancing poetic design with technical constraints. Technical planning is what allows that richness to survive across platforms and devices. When it works, players do not see the optimization behind the scenes. They simply feel that the world is full, natural, and alive.

For me, the most important outcome was that performance planning did not flatten the artistic intention. It allowed the city to keep its narrative density while still functioning as a playable live-service environment.

Game Art as Designed Experience

In a standalone visual project, the artist is often shaping a controlled expression: a final image, shot, or sequence where the viewing experience is largely decided. Game art asks for a different kind of authorship. Once an environment becomes playable, the artist gives up part of that control. The space has to remain coherent when players move through it at their own pace, view it from unexpected angles, find their own focus, or return to it with a different purpose.

That is why I approach game environments less as finished images and more as designed conditions for experience. I still care deeply about composition, lighting, silhouette, and mood, but I also have to think about how the space behaves once it is in the player's hands. The environment needs to give players a clear sense of direction, feel right in relation to the avatar, and remain visually coherent even when the camera is not perfectly controlled. The craft is in building a space that can guide players naturally without making the experience feel overly directed.

The more difficult part is that all of this has to work in real time. In a rendered project, complexity can often be absorbed into production or render time, but in a game, it becomes part of the player's experience. A beautiful environment still has to run smoothly, support movement, and remain clear while the player is moving through it. That changes the way I make visual decisions. I judge each element by what it contributes to the experience: areas that carry gameplay or emotion deserve more detail, while supporting areas can stay more suggestive so the world feels rich without becoming visually or technically overwhelming.

The most exciting part is that game art leaves room for the player to complete the work emotionally. In Sky, the same environment can become a very different memory depending on the player's mood, their companions, and the way they choose to spend time there. A space may be designed with a clear function, but players often give it meanings the artist could not fully predict.

That variability is not a weakness of the medium; it is part of its beauty. For me, creating art for games means building a world that is intentional enough to guide the player, but open enough for them to feel ownership over the experience. The artist shapes the world and the conditions for discovery, but the meaning of the experience is ultimately completed by the player.

Advice for Artists Entering Games

For artists who want to enter the game industry, I would first encourage them to build a strong foundation before trying to specialize too quickly. Early in your career, it is very helpful to understand the broader 3D production pipeline: modeling, lighting, materials, composition, basic animation, and how assets are prepared for production. You do not have to be an expert in everything, but having that generalist foundation gives you better judgment later. Over time, you will naturally discover which part of game development you are most drawn to and where you want to go deeper.

For artists coming from other areas of 3D production, I think learning at least one game engine is one of the most important steps. Build a small playable scene from beginning to end. It does not need to be huge, but it should let you experience what happens when your work becomes a space that someone can actually move through. That process helps you understand how game art is different from a render, animation, or still environment piece. You begin to think about the player's movement, point of view, pacing, scale, readability, and how the environment communicates with a free camera.

I also think it is important to test your own work with other people. As artists, we often know our own intentions too well, so we may not see where the experience is unclear. Let someone else play through the scene and observe what they notice, where they go, and how they describe the mood or story afterward. Then compare that with what you intended. If the emotion or message is not coming across, that is valuable information. Learning how to test, receive feedback, and iterate is one of the best ways to start thinking like a game artist.

Collaboration is another skill I would take seriously. Games are made by teams, and a lot of production depends on clear communication with designers, engineers, animators, other artists, and leads. Good communication can save a huge amount of time. It is also important to become comfortable with critique. Being able to give useful feedback, receive feedback without taking it personally, and discuss artwork in a way that serves the project is a big part of working professionally.

For actually breaking into the industry, I would recommend attending game conventions, portfolio reviews, and industry events whenever possible. Pay attention to the studios and games you admire, and see where their developers are speaking, exhibiting, or reviewing portfolios. Do not be afraid to reach out, introduce yourself, and show your work. Many internship or junior opportunities can start from a portfolio review, a short conversation, or a simple follow-up email. Strong work is important, but being able to explain your process and show genuine interest in the industry can make a real difference.

Overall, my advice is to stay curious and keep building. Learn the tools, make playable work, ask for critique, meet people in the industry, and gradually find the area of game development that feels most meaningful to you. The industry can feel difficult to enter from the outside, but every small project, conversation, and iteration helps you understand it better.

Flora Yu, 3D Artist