Hair in Video Games: How and What Technologies Were Used to Create It

Contents:

• 1990s: Polygonal Meshes and the First Steps in Simulation
• 2000s: New Shading Possibilities
• 2010s: The Pursuit of Realism
• Present Day and the Future

An article in Vice magazine noted that the evolution of 3D graphics in video games can be traced through the hairstyles of characters. With the development of technology, this element of the image is becoming increasingly realistic. However, creating hair in games, like facial animation, remains a challenge for developers and artists, especially when striving for realistic graphics. Issues of hair physics, lighting, and interaction with the environment require significant effort and innovative solutions. Effective implementation of these aspects contributes to the overall quality of the game and creates a deeper immersion in the gameplay.

To understand why creating hair in video games is a complex task, it is important to go back to the origins and examine the evolution of technology in this field. Developing realistic hair requires taking into account many factors, including physics, textures, and animation. Over time, technology has improved, but challenges remain. For example, realistic hair modeling requires significant computational resources, making the task even more challenging for developers. Furthermore, the characteristics of different hair types and their interaction with light add additional layers of complexity. Therefore, creating realistic hair in games remains an important and interesting topic that requires constant improvement and innovation.

1990s: Polygonal Meshes and the First Steps in Simulation

The first attempts to create unique hairstyles for characters appeared in 3D games of the early 1990s. In these games, the characters' hair was represented by a set of polygons and was an important part of the character model. Textures were used to separate the boundaries of the hairstyles from the overall polygonal mesh. Even on low-poly models, artists strove to create elements resembling hairstyles, from ponytails to buzz cuts, as, for example, in the 1993 game Virtua Fighter. These early experiments laid the foundation for the subsequent development of video game graphics, where character customization, including hairstyles, became increasingly important.

If a character's design requires long hair, this creates additional challenges. First, there's the risk of exceeding the allowed polygon count, which can negatively impact performance. For example, in the first Tomb Raider, due to technical limitations related to collision and the polygon limit, Lara Croft's long hair was replaced with a bun. This solution illustrates how the need for graphics optimization can impact character design in video games.

Long hair, unlike short haircuts and beards, presented certain challenges in the solid polygon mesh format. In such cases, the mesh was attached to the character's head, which resulted in unnatural texture stretching when the neck turned. Furthermore, if the hair was pulled back into a long ponytail or braid, additional animation development was necessary. This required additional resources and effort to achieve natural movement, which is essential for creating a realistic gaming experience.

In the original Tomb Raider, character animation was created not using a skeletal rig, as is common in modern games, but based on joints. Joints were small fragments of a polygon mesh that controlled the rotation of individual body parts. By programming the joints' movement, the developers could create various poses for the heroine, creating animation cycles based on them. With the release of Tomb Raider 2, Core Design managed to implement the scythe into the game as a separate mesh, which significantly improved the visual perception of the character and animation.

Lara's hairstyle was created using multiple joints, allowing for realistic animation synchronized with the character's movements. For example, when Lara fell, her braid would rise, and when she jumped horizontally, it would swing back. The dynamics of the hairstyle's joints also depended on triggers that responded to environmental conditions. In open spaces, the braid could sway in the wind, while underwater, it would twist and gradually rise. This approach was innovative for its time, given the generally static nature of hairstyles in video games. The introduction of such technologies made gameplay interactions more lively and engaging, increasing player interest in the visual aspects of the characters.

In the mid-1990s, the first experiments in motion simulation in video games began. One striking example is Virtua Fighter 3, which utilized physics to create realistic animations. Particularly notable are the movements of characters' long hair, such as Aoi Umenokoji's ponytail, demonstrating the high degree of physics-based gameplay. This development was an important step in the evolution of graphics and animation in video games, opening up new possibilities for creating more lively and dynamic characters.

In Drakan: Order of the Flame, protagonist Rynn's long ponytail sways dynamically as she runs, creating a realistic motion effect. When she jumps, her ponytail lifts slightly, adding atmosphere and visual flair to the gameplay. These details highlight the sophistication of the character's animation and help immerse players in the game world.

However, the examples above are unique. As mentioned earlier, game developers generally avoid long hair. If long hair is present in games, it most often has limited dynamics and is controlled by pre-defined animations. This is due to technical limitations and the need to optimize performance, making the implementation of realistic hair movement a challenging task.

Technical limitations forced developers and designers to focus on gameplay mechanics and environment creation, which led to the fact that the detail of the characters' appearance was often compensated for by textures and spectacular skins. This explains why Julia's appearance from the animated film "Heavy Metal 2000" was changed in the game Heavy Metal F.A.K.K. 2, and her luxurious hair was replaced with a simpler bun. This approach allowed us to maintain a high quality of gameplay, even if the visual characteristics of the characters were simplified.

If the game features characters with long, flowing hair, the hair becomes an integral part of the model's polygonal mesh, flowing smoothly into the back. In cases where hair is extremely long, texture rendering is used to achieve a realistic effect. This allows you to create more detailed and attractive images, which helps improve the visual perception of the game and increases interest in the characters.

In the 1990s and early 2000s, a unique approach was used to creating hairstyles. Hair was often shortened or shaped into simple geometric designs. Textures were used to visualize fur or stubble, which added realism. This style became iconic for its time and influenced subsequent trends in character design.

2000s: New Shading Possibilities

The situation changed with the development of game engines, which began to support a higher number of polygons in scenes and shaders with transparency effects. The basic structure of hair meshes remained inseparable from the character model. However, the introduction of transparency maps gave 3D artists the ability to create more varied and realistic hair variations. This innovation significantly improved the visual perception of characters and expanded the creative possibilities in the gaming industry.

The image below shows Yuna's model from the game Final Fantasy X, released in 2001. Note the structure of her hair, created using a ladder technique with textures with a transparent effect at the ends. The translucent ends and thinning make Yuna's hairstyle appear more natural and realistic. This detailed hair work highlights the high quality of graphics and attention to detail characteristic of the Final Fantasy series.

In Silent Hill 2, the hair modeling technique is implemented differently. For example, the ends of Maria's hair are ribbon-like planes that use shaders to create a transparent effect. A key feature is the ability to dynamically displace polygonal strands, which adds realism. This method is also effective for short haircuts, allowing the bangs to be emphasized. This approach to hair rendering helps create depth and expressiveness in characters, which is an important aspect of the game's atmosphere.

The collage shown shows that the model has separate eyelash meshes, which are created using slightly convex planes. This approach allows us to preserve the natural curl of the eyelashes and ensures their realistic appearance.

The image in which the hair reaches a length below the shoulders is often complemented by ribbons of polygons, visually lengthening the hairstyle. A prime example of this is Claudia's model from Silent Hill 3, where the use of such elements emphasizes the character's personality and atmosphere.

Metal Gear Solid 2: Sons of Liberty (2001) features an interesting design approach for Raiden: his hairstyle is represented as a separate mesh that replaces part of his head. This approach allows for more realistic hair movement during dynamic cutscenes. Furthermore, this approach is connected to a unique gameplay element—the ability to equip Raiden with bonus wigs of various colors. These wigs not only change the character's appearance but also possess unique characteristics that influence gameplay. This adds an element of variety and strategy to the game, allowing players to tailor their playstyle to their preferences.

Transparency maps have significantly improved the realism of character hair in games. However, the challenge of conveying the natural dynamics of hair as it moves remains.

The engine of The Legend of Zelda: The Wind Waker, released in 2002, introduced innovative technologies such as clothing simulation, flexible objects including ropes, and realistic hair movement. Thanks to these advancements, even Link's stylized, minimalist bangs responded smoothly to his movements, adding depth and dynamism to the gameplay. These elements not only improved the visual experience but also made the gaming experience more immersive and realistic for players.

In the context of creating animals and creatures, developers often use not only the technique of working with planes, but also special shaders that imitate the fur texture. While the polygonal model structure remains unchanged, the key feature is the use of innovative rendering technologies. These shaders allow for realistic fur rendering, adding depth and detail that significantly improves the visual appeal of models. The use of such technologies is a key step in creating high-quality graphics in video games and animation.

The AtmosFear engine, used in the shooter "Vivisector: The Beast Within," has become renowned for its ability to realistically render animal fur and anthropomorphic creatures. This game stands out from the crowd thanks to its high-quality graphics, which create a realistic and immersive atmosphere. Visual effects, implemented using AtmosFear, significantly enhance player immersion in the gameplay and make it more exciting.

Information on how Action Forms achieved such an impressive effect is very limited. The media landscape of the time was dominated by only general enthusiastic comments. Considering that the developers used the high-level language HLSL for shader programming, and that each animal model in the game texture files has an additional mask with a semi-transparent layer, it can be concluded that the pixel fur shader was developed based on complex calculations. This approach significantly improved visual effects and created more realistic graphics, which, in turn, attracted the attention of players and critics to the project.

A shader "recipe" can be found online, which is similar to the one used in the game "Vivisector." Comparing online examples with models from this game, one can conclude that the developers used a similar technique. While some users felt that the visual rendering of fur in Vivisectionist was inferior to the animal fur in Black and White 2, released during the same period, the shaders used in Vivisectionist were still quite impressive for 2005.

In Shadow of the Colossus, the fur of the giant creatures was created using a fur shader, but without the need for programming. Several parallel planes with varying transparency levels are superimposed on the main model. This creates a "spike" effect from randomly placed areas of the fur texture, allowing for the formation of a three-dimensional texture on the skin. This approach ensures realistic depiction of animals and adds depth to the visual perception of the game, which emphasizes its artistic value and technological achievements in the field of graphics.

The approach taken by the developers of Shadow of the Colossus remains relevant in modern games. A prime example is Stray, released in 2022, which uses similar mechanics and design elements. This connection between the projects highlights the impact that Shadow of the Colossus has had on the gaming industry and the inspiration it has provided to new developers.

Let's return to the topic of hair. In the era of the seventh generation of gaming consoles, such as the PlayStation 3 and Xbox 360, some games began developing hairstyles as separate meshes, similar to Raiden's model discussed earlier. This allowed developers to assign unique physical properties to different parts of the hair, which significantly improved the realism of their animation and interaction with the environment. Thanks to this step, the gaming industry was able to come closer to creating more lively and dynamic characters, which affected the overall level of immersion and perception of games.

In 2007, the slasher Heavenly Sword was released, using an advanced version of the Havok physics engine. This technology played a key role in simulating the hair of the main character, Nariko. Although the long red polygonal strands sometimes looked a bit chaotic in dynamic mode, the bold fantasy look became a distinctive feature of the character. Heavenly Sword received acclaim for its innovative graphics and immersive gameplay, making it a landmark title in the video game industry.

By the end of the decade, artists began to actively use the technique of mapping small planes with translucent textures. These textures were created based on modeled hair curves that were baked into map sets. Known as Hair Cards or Hair Sheets, these maps were applied to small planes, allowing for realistic detail. This resulted in polygonal fragments being evenly distributed across the base model, significantly improving the visual quality and believability of the depicted objects. This technique became an important step in the development of high-quality 3D graphics and animation.

Gears of War 2 was one of the first games to employ the technique of using translucent textures to create realistic effects. Clusters of planes with such textures not only added visual volume to hairstyles but also provided a more realistic display of facial hair. Just ten years ago, beards and stubble were created using textures or additional relief of the polygonal mesh, which significantly limited the visual capabilities of games. Modern technologies allow for a new level of detail and realism, which makes the gaming process more engaging and immersive.

The Hair Cards or Hair Sheets technique is based on principles similar to creating eyelashes and shaping strands of hair. However, the process requires greater care, as the length of individual "pieces" can vary, as well as their texture. This creates a more natural-looking base, as the strands are formed using curves rather than created manually or obtained through photoscanning. This allows for a higher level of realism in hair images, making the technique particularly popular in 3D modeling and animation.

The range of plane applications has expanded significantly, and artists have begun using them to shape eyebrows. This method has become popular due to its precision and the ability to create a natural look. The use of planes in cosmetic art helps achieve perfect symmetry and highlight individual facial features.

2010s: The Pursuit of Realism

In the 2010s, game developers significantly increased their focus on the realism of character hair. The polygonal mesh of the hairstyle evolved into a separate element, resembling a wig, created from numerous Hair Cards. This approach allowed for the creation of more detailed and natural-looking hair, which ultimately improved the visual perception of game characters and increased the level of immersion. The use of Hair Cards has become an industry standard, allowing artists and designers to achieve impressive graphical results.

This method not only provided the necessary level of realism but also provided the ability to quickly replace one mesh with another when a character's appearance changed during the game. This approach also proved effective in creating NPCs, allowing the reuse of the same elements. Today, forming hair from individual meshes or mesh groups is considered a common practice in the gaming industry.

There are many methods for creating polygonal hair structures in video games. The choice of approach depends on the visual style of the game and the characteristics of the character's hairstyle. Most often, artists use two main methods: sculpting followed by texture mapping or shaping hair using specialized software that supports simulations and procedural curve-based generation. Splines are created in areas where hair is expected to be present, which are then transformed into curves from specified points. Once the desired shape is achieved, the design is optimized, and depending on the model's complexity, the result can be baked into one or more meshes. This process achieves a high degree of realism and detail, which is critical for modern video games. It's worth noting that creating hair using procedural geometry and real-time particle systems, as implemented in Blender or the Yeti plugin for Maya, produces impressive results. However, this method is resource-intensive, a critical factor in video game development. The simulation process involves constant calculations of physics, collisions with other objects, light reflection, and other characteristics of each individual hair. Despite its high performance requirements, this technique is widely used in animated films, where visual quality is crucial. Facial and body hair, including eyelashes, eyebrows, stubble, as well as chest and arm hair, have come to be classified into separate mesh groups. An example of this is the basic mesh shapes created using Hair Cards, which were used in games in the early 2010s. This approach allows for more accurate display of hair details and variety in game projects, which significantly improves the visual perception of characters.

The method of generating hairstyles using planes and translucent textures remains effective in terms of performance. However, with the development of technology and the increased realism of graphics in video games, new challenges have arisen in creating believable hair. Modern games require more detailed hair models, which poses the challenge of finding a balance between graphic quality and performance.

A new rendering technology was soon introduced that significantly improved the display of hair in 3D. This technology, known in the industry as the strand-based method, is based on the concept that the word "strand" translates as "thread," and in this context, as "hair." The basic principle is that a hairstyle on the screen is converted into thousands, sometimes millions, of curved strands, which represent virtual polygons. This process is typically performed using the Alembic graphics framework. The visual characteristics and physical properties of such hair are as close to real hair as possible, achieving a high degree of realism in computer graphics. This innovation opens up new horizons for character and animation creation, making them more natural and engaging for viewers.

The main drawback of this method is the high computational cost of rasterization, which requires a powerful graphics card. However, AMD's TressFx and NVIDIA's HairWorks technologies, which appeared in the 2010s, successfully attracted the attention of players thanks to realistic graphics for their time. This also became an incentive for developers to pay attention to new approaches in graphics creation.

TressFx gained popularity thanks to its integration into the reimagined Tomb Raider, released in 2013. Lara Croft's new hairstyle not only displayed every hair in detail but also demonstrated realistic physics, reacting to the character's movements and environmental conditions such as water, snow, and wind. Currently, the fifth version of TressFx technology is available, which can be integrated into Unreal Engine, but so far only for versions 4.26 and 4.27. TressFx continues to be an essential tool for game developers looking to create more realistic visuals in their projects.

HairWorks is an advanced technology from NVIDIA that combines previously developed methods for rendering and simulating hair, fur, and wool. HairWorks plugins give artists the ability to customize the density and shape of hair in 3D modeling programs. This technology significantly improves the realism and quality of rendering, allowing you to create detailed and believable images of characters and environments. Using HairWorks in projects contributes to an overall higher level of graphics and draws attention to small details, which is especially important in modern video games and animation.

HairWorks technology was implemented in many games of the 2010s, including Far Cry 4, Call of Duty: Ghosts, and The Witcher 3: Wild Hunt. In The Witcher 3, users often expressed dissatisfaction with the incorrect rendering of Geralt's hair, which led to the creation of a special mod to fix this issue. This mod became popular among players looking to improve the visual quality of the game and increase the realism of the character.

In the 2010s, various hair rendering and simulation technologies were developed in the industry, which were used in various software products and game engines. Many of these technologies remain relevant today, continuing to influence modern hair modeling in 3D graphics.

• The Dawn Engine from Eidos Montréal had its own system for rendering character hair, which was essentially an improved version of TressFx technology. An example of its use can be found in Deus Ex: Mankind Divided.
• The Luminous Engine (Square Enix games) had its own hair and fur rendering technology. By the end of the decade, it became known as Luminous Hair.
• The Frostbite engine's built-in hair system (Mass Effect: Andromeda, Dragon Age: Inquisition). Although Frostbite's approach was inferior to other solutions in the industry, by the end of the decade the technology had evolved significantly, and the result of this evolution can be seen in the video below. We will discuss the specifics of this approach below.

• The Unreal Engine supported improved hair shaders as early as its fourth iteration. Towards the end of the decade, it introduced its own Groom system for creating realistic hair simulations using the previously mentioned strand-based method.
• A procedural method for creating hair and fur in Houdini.
• Various plugins for software and engines. Among the most famous are XGen from Maya; Ornatix for 3ds Max and the NeoFur plugin for Unity by NeoGlyphic (support for this technology is currently discontinued).

By the end of the decade, game engine and hardware technologies had reached a level that allowed games to display high-polygon models and high-resolution textures. This opened new horizons for artists, who began to implement the most complex and ambitious ideas. A prime example is Aloy's hairstyle from Horizon: Zero Dawn, which has become her signature look. This hairstyle is made up of 100,000 triangles and animated using 50 splines, demonstrating the level of detail and polish achieved in modern video games.

Present Day and Future

Modern developers continue to strive for excellence in hair rendering in video games, refining existing technologies. In most AAA games, artists use a comprehensive approach to creating hair, combining the Hair Cards technique with strand-based rendering. This allows for a powerful graphics card to achieve detailed hair, where each strand is visible up close. At low detail settings or on less powerful systems, players see hair composed of polygonal strands and translucent shaders. This approach provides a balance between graphic quality and performance, making it ideal for modern gaming projects.

When working with stylized characters, 3D artists continue to use sculpting to create hairstyles. The main changes concern improved physics and the use of shaders, which provide more natural and softer light reflection. Such technologies allow us to achieve a high level of realism and artistic expressiveness in 3D modeling, which is especially important for creating memorable and attractive characters.

The industry of realistic hair rendering continues to develop actively, and research in this area does not cease. Among the most recent achievements, several innovative technologies stand out that have significantly improved the quality and realism of hair rendering. These innovations open new horizons in the creation of visual effects and animation, allowing developers to achieve impressive results in their projects.

• The Groom plugin in Unreal Engine 5. Supports both the import of strand-based hair systems in the format of the special Alembic framework, and alternative hairstyle geometry options in the form of maps and meshes. For more precise hair adjustments, the built-in Groom Asset Editor is available.

• Unity's hair rendering and simulation system, previously featured in the Enemies tech demo. In the future, an improved version of this technology will appear in the stable version of the Unity 6 engine.

• The GroomBear toolset for Houdini. Contains specialized nodes for generating fur and hair within the software, brushes, and nodes for converting to geometry (e.g., for planes, feathers, etc.).

• An improved technology of the Frostbite engine mentioned above. The new system is also based on the strand-based method, and the hair dynamics are determined by physics. The results of this progress can be assessed not only by the hairstyles of football stars in modern simulators like EA Sports FC, but also in the recently released trailer for Dragon Age: The Veilguard, which surprised players with the realistic physics of long hair and its display in real time.

The new technology deserves a more detailed examination. Previously, Dragon Age fans often expressed dissatisfaction with the technical limitations of the series, which did not allow the creation of characters with long hair. The implementation of hairstyle dynamics and their appearance also did not meet player expectations. However, BioWare took community feedback into account and took a significant step forward by providing more realistic rendering of character hair. Players can now enjoy improved graphics and more detailed hairstyles, making the gaming experience more engaging and immersive.

A new approach to hair processing in the Frostbite game engine involves importing hair in NURBS format. This method creates a tessellation along each strand, forming a set of points to construct a curve. The more points used, the more realistic the image becomes. However, to achieve an optimal balance between realism and performance, the developers limit the number of points. In the Frostbite demo video published earlier, 25 points were used per strand. The use of NURBS in hair creation is expected to become more widespread in the future, which will improve the quality of graphics in games.

In recent articles, we discussed how 3D graphics professionals are beginning to actively use neural networks. Hair creation in this context is no exception. Among modern developments, it is worth noting such projects as Digital Salon and Haar. NVIDIA is also actively researching generative artificial intelligence and neural graphics, involving dozens of scientists from various institutes around the world in their research. This demonstrates growing interest in the application of advanced technologies in 3D modeling and the creation of realistic visual effects.

In the future, the results of this research will have a significant impact on professionals in fields such as art, architecture, graphic design, game development, and film. These advances will help speed up the process of creating high-quality content necessary for storyboarding and previsualization, and can also be used in game development. Optimized workflows and increased efficiency will significantly improve the final product in these creative industries.

Creating realistic 3D hairstyles is a complex task, primarily due to the need to accurately reproduce the dynamics inherent in real-life hair. Hair movement depends on many factors, including wind, head movement, and the physical properties of the hair itself. Furthermore, even with the most modern technologies in 3D modeling and game engines, the implementation of these movements is limited to certain scenarios and numerous settings. Effectively conveying natural hair dynamics requires a deep understanding of both physics and software features.

The situation is further complicated by the fact that hair characteristics must change under the influence of the external environment, including physical properties when interacting with water, dirt, or sunlight. However, in this regard, hair realism remains insufficient. For example, Unreal Engine often has problems with correctly displaying hair color due to lighting conditions. This highlights the need for improved rendering technologies to achieve a more natural and realistic visual representation of hair in gaming and multimedia applications.

Modern game worlds containing numerous NPCs require careful optimization, which directly impacts rendering quality and the level of detail. Optimizing gameplay ensures smooth operation and improves the user experience, which is especially important in open-world games. Effective optimization methods help developers achieve a balance between visual richness and performance, making the gaming experience more engaging and immersive for users.

Despite advances in graphics and animation, the uncanny valley effect still persists. This problem arises due to imperfect rendering and physics technologies, which hinder the creation of fully realistic character hair in games. Developers are actively working to improve these technologies, striving to achieve the highest level of realism and eliminate discomfort associated with the perception of virtual characters. As a result of continued efforts in this area, we can expect significant breakthroughs in the future, leading to more natural-looking hair and other aspects of game characters' appearance.

Modern graphics standards are becoming increasingly blurred, due to the popularity of various stylizations such as cel-shading and low-poly. In the gaming industry, the rendering of hair on characters does not always achieve a high level of detail. In some indie projects, designers may use simplified modeling techniques, following a specific concept. However, should we blame them for this? History shows that such experiments can lead not only to solutions to current problems but also to the discovery of new aspects of virtual hairdressing. It is important to understand that innovations in graphics can arise precisely from non-standard approaches, which makes each game unique and attracts the attention of players.