Third Person Camera View in Games - a record of the most common problems in modern games, solutions taken from new and retro games

The aim of this article is to provide an overview of the most common camera problems of third-person games. For this purpose, a comparative analysis of various 3D games that utilise the third-person view perspective is carried out. In addition, solutions and workarounds will be presented which serve the development of a robust camera system.

First, the basics of a virtual camera system are elaborated. In addition, the different behaviours of cameras are presented. A comparative analysis of several third-person games will list the most common problems and particularly positive examples. Finally, solutions will be presented which can be used for the development of a camera system.

Chapter 1: Basics of Game Camera Views and Systems

In order for a player to participate and act in a video game, he must be able to perceive the environment of the video game (Thorn, 2013, p. 10). He must be able to see what is happening and react accordingly. This game world is communicated to the player through the lens of a virtual camera located at a certain position and oriented in a certain way in the environment. 

1.1 Camera Systems

The term camera system includes various components combined to allow the video game to control how the virtual world is presented to the player (Haigh-Hutchinson, 2009, p. 28). It is a key component inside the game engine and performs the following tasks, among others:

1.2 Camera Behaviours

By camera behaviour one means the combination of camera characteristics (e.g. required position, field of view, movement characteristics, etc.) which controls the displayed game world (Haigh-Hutchinson, 2009, p. 29).

1.2.1 What is a First-Person View?

In the First-Person View, the camera is located inside the eyes of the main character (Reeves & Read, 2009, p. 97). You can't see the face of your own main character, but in certain cases body parts can be visible (e.g. the feet of the main character if you look down).

Fig. 1: Halo 5: Guardians - First-Person View Example

The first-person view gives the player the feeling of being "in the shoes" of the main character (Chacon, 2017, p. 43). This makes a video game feel more real. The disadvantage of this camera style is the limited field of vision, similar to real life. The player only sees a very limited part of the 3D landscape and often doesn't have the possibility to switch to different perspectives.

1.2.2 What is a Second-Person View?

The second-person view describes the view of a main character through the eyes of another character (Perry & DeMaria, 2009, p. 33). A good example of such a second-person view can be found in the video game Battletoads (Rare Ltd., 1991), which was released in 1991 for the Nintendo Entertainment System (Douglass, 2007, p. 147). Even before you see your opponent in a fight, the view changes to the second person view. This means that you see the fight from the opponent's point of view. But you still control the main character, so the player has to aim the attacks at the screen he looks through himself. The following picture illustrates this view:

Fig. 2: Battletoads - Example of a Second-Person View

1.2.3 What is a Third-Person View?

A camera behavior in which the camera is separated from the focus of the main character is called a third person view (Haigh-Hutchinson, 2009, p. 52). Sometimes, this camera behavior is also called a bird’s eye view, since the game world is not viewed from the eyes of the main character, but from the perspective of an external position. The weaknesses of the first-person camera are the strengths of the third-person camera: In contrast to the first-person view, the player has a much wider field of view with this camera behavior, he can see his main character as a whole and can view the 3D landscape from different perspectives (Chacon, 2017, p. 43f.). If a fight takes place in the video game, the third-person view offers the player a better overview, since he can see the whole environment, including which opponents are behind or next to him. So he can defend himself better. 

In contrast to the first-person view, the third-person view allows the relationship between the main character and the environment to be better demonstrated (Haigh-Hutchinson, 2009, p. 52). This allows more intense interactions in games where the game world is an important part of the gameplay. Super Mario 64 (Nintendo EAD, 1996), a game of the "3D Platformer" genre, is one of the earliest games to use this view, and is now considered a prime example of third-person viewing. Particularly in this game, the camera was integrated into the game as a character: Lakitu, a character who is visible from time to time, follows Mario during the game (Nitsche, 2008, p. 96). The game allowed the player to control Lakitu's camera, allowing Mario and the game world to be viewed closely. Such games require that the main character and the camera be controlled simultaneously by the player (Nitsche, 2008, p. 97). 

Fig. 3: Super Mario 64 - Third-Person View Example

A third person camera has an enormous influence on the visual representation in the game. If the camera is completely controlled by the player and the game world can therefore be explored exactly, the game world is no longer just the "background" of an image (Nitsche, 2008, p. 97).

The third-person camera is often used in certain situations in games that mainly use the first-person view (Adams, 2010, p. 412). For example, in Halo (Bungie Studios, 2001), the view changes from first-person to third-person when the main character enters a vehicle. This triggers a change in camera behavior as the vehicle temporarily becomes the main character controlled by the player.

The implementation of a third-person camera can be more complex from a game design perspective than the implementation of a first-person camera (Chacon, 2017, p. 44). Since you don't see the main character of the first-person camera, but only the hands and weapons, you also need less character animations. The third person camera, on the other hand, requires a complete animation of the character, e.g. for running, crawling, jumping, etc.

Chapter 2: Third-Person Camera Positions

Third-Person Camera Views aren’t always in the same place. In this chapter the various ways of positioning the camera will be showcased and explained thoroughly.

2.1 Position of the Camera in the Video Game Environment

By definition, third person cameras are not "in the eyes" of the main character, but in an external position, "outside" of the character (Haigh-Hutchinson, 2009, p. 219). In many si-tuations it is necessary that the camera position takes over the viewing direction of the main character. This allows an unobstructed view of the game world. The following two camera positions are common in practice:

Fig. 4: Camera behind and above the main character, centered

Fig. 5: Camera at shoulder height of the main character, not centered, but slightly to the left or right of it

There are several reasons for centering a third person camera perspective. If a character (or an object) is centered within the composition, i.e. in the center of the representation, it is given a central meaning (Solarski, 2017, p.30). The positioning of the camera thus has a major influence on the narrative of the story. When choosing the camera positioning, visual and narrative aspects have to be taken into account.

2.2 Variants of Camera Control

There are several ways to control the camera. The variants of the camera controls are first divided into three categories: Automated, Player Control and Hybrid.

2.3 Automated

There are some video games where the entire control of the camera is taken over by the programming of the game, making it impossible for the player to control the camera (Haigh-Hutchinson, 2009, p. 219). This may be necessary to ensure that the camera remains in the game world, but in most cases the goal is to save the player the effort of controlling the camera. 

Since the course of the automated camera can be controlled, the camera allows to set a special focus in a certain situation, e.g. on a certain object or target. 

The disadvantages of an automated camera control are that the player may feel impaired and that great care is required during game development so that the camera can display all possible gameplay situations (Haigh-Hutchinson, 2009, p. 219). 

2.4 Player Control

Third person games often allow the player to control the camera to a certain degree (Haigh-Hutchinson, 2009, p. 219). However, there are often restrictions on how much con-trol the player has. 

In racing games, predefined views of camera alternatives often exist, e.g. the one from within the car (Haigh-Hutchinson, 2009, p. 219). The player can then only move the came-ra slightly within this predefined view. 

Video games in the genre of "massively multi-player online games" (MMOG), on the other hand, allow the player full control over the camera, both while strolling through the game world and outside the game world (Haigh-Hutchinson, 2009, p. 219). However, such great freedom with regard to camera control is not recommended in the literature. There is a risk that the game world will be represented inappropriately.

2.5 Hybrid

With the hybrid version, the majority of the camera control is automated. However, the player is still given the option of controlling the camera himself to a limited extent (Haigh-Hutchinson, 2009, p. 220). 

For example, a hybrid variant is the following: The player has the possibility to rotate the camera to perceive the surroundings (Haigh-Hutchinson, 2009, p. 220). As soon as the player stops using the camera rotation, the camera positions itself again at the programmed location. 

Chapter 3: Third-Person Cameras in Comparison 

The following chapter shows the most common problems of third-person cameras based on a comparative analysis.

3.1 Central Problem Fields

The comparison of different video games shows that there are many problems and difficulties with regard to third person cameras. The main problems can be summarized as follows.

3.1.1 Non-Centered Camera

In Elder Scrolls Online (ZeniMax Online Studios, 2014) the third person camera (Player Control variant) is located at shoulder level of the main character. It is slightly offset to the right. In the middle of the screen there is a targeting disk, which represents the center of the camera view, and not the center from view of the main character.

If the player wants to run through a goal, he aims the disc at the open column. Since the camera is offset, the main character will not pass through properly, but will either run into the door and get stuck, or the camera will bounce off the other door, which could frighten the player.

Fig. 6: Elder Scrolls Online – Targeting Disk Offset

During a fight, the player targets an opponent by means of the targeting disc, but he will miss him, because the discis the center of the camera view, but not the center from the point of view of the main character. The camera is not centered on the character. The player is therefore forced to change the position of the main character so that it looks at the opponent, even if the target is to the right of the opponent. 

Fig. 7: Elder Scrolls Online – Targeting Disc in a Fight

In The Last Guardian (genDESIGN and SIE Japan Studio, 2016) there are two main characters, only one of which can be controlled directly by the player. The secondary character follows the main character and listens to his commands, such as attacking a target. It is much larger than the main character and does not fit into the full camera view of the main character. The camera follows the main character as he walks. However, when the player stands still and wants to look at the surroundings, the camera often focuses on the secondary character. It is therefore a variant of the hybrid camera control system. This prevents the player from exploring the game world and is forced to constantly adjust the camera.

3.1.2 Camera is Covered by Objects

A very common problem is that the camera is covered by objects. An example is the third person camera in Final Fantasy XV (Square Enix Business Division 2, 2016). This is designed at shoulder height of the main character and can be controlled by the player (Player Control). The environment in this video game is rich in flora and fauna. The problem is that the camera is often hidden by trees, leaves and plants. The player must constantly ad-just the camera himself to improve his viewof the game world and to understand what is going on. This problem of the hidden camera can be very annoying for a player. In particu-lar, the game Final Fantasy XV offers a very large, open game world (Open-world). The ex-ploration of this world is a big, important part of the gameplay. If the camera is constantly covered by the trees and bushes of the landscape, this disturbs the player. Furthermore, it can happen that the camera is hidden in the middle of a fight against one (or more) oppo-nents, so that the player can no longer see what is actually happening. This way the player cannot defend himself properly and cannot control his attacks properly.

Fig. 8: Final Fantasy XV - Camera is Hidden

3.1.3 Camera bounces off Objects

Assassin's Creed Syndicate (Ubisoft Quebec, 2015) has a third person camera as player control variant, which is located behind and above the main character. The video game takes place in London, with many non-playable characters (NPC's) and narrow alleys. As a result, the camera often bounces off objects and walls, which can irritate the player.

Dark Souls 3 (From Software, 2016) has the same third-person camera, which is much too close to the main character. In this game there are often fights with huge opponents. The camera often clings to the legs of these enemies or bounces off the walls in tight environments. This can make these fights a lot more difficult, which wasn't even conceived during the development of the game. 

Cameras often have a "Sphere Collider", which is often the cause of this problem. The following figure shows how a "Sphere Collider" works:

Fig. 9: Sphere Collider – Unreal Engine 4

A Sphere Collider is an implicitly defined and invisible sphere (Ericson, 2004, p. 10). Schematically it can be represented as a virtual sphere with an interior space and outline that bounces off other objects, provided they also have a collider. 

3.1.4 Risk of Kinetosis

Kinetosis or nausea is also called travel sickness because it occurs mainly when travelling by ship or plane (Stoffregen, Yoshida, Villard, Scibora, 2010, p. 169). However, it can also occur with gamers. There are various theories on the causes of kinetosis, but most of them assume that the cause is an intersensory conflict (Stoffregen, Yoshida, Villard, Scibora, 2010, p. 170). The messages of the different sensors do not fit together (Scherer, 2013, p. 175). The player hardly moves himself, but he looks at the game in which he moves his main character. Other theories dealing with kinetosis in VR games focus on the fact that movement control is limited (Riccio & Stoffregen, 1991, p. 195). 

Gears of War 4 (Black Tusk Studios, 2016) has a third-person camera, which is located at shoulder level of the main character and can be controlled by the player (Player Control). Although the camera is not in the eyes of the main character, the camera wobbles during running and sprinting, as is often the case in first-person games. The connection between the character animation and the camera shake gives the player a sense of realism and aggression (Solarski, 2017, p. 21). This can lead to the player becoming nauseous due to the constant movement of the camera (GDC Vault, 2014). 

The cause of nausea when playing are the camera movements. The developer Tsurumi Roppyaku emphasizes that especially the lateral camera rotations and the way in which the camera follows the main character are difficult (Isbister & Schaffer, 2011, p. 171). In his opinion, a camera that goes up and down and rotates is particularly problematic. Furthermore, too fast or too slow a camera movement can lead to nausea.

Fig. 10: Gears of War 4 (Camera Shake)

Chapter 4: Solutions for a robust Third-Person Camera View-System

This chapter presents, based on a comparative analysis, solutions to the problems listed in the previous chapter.

4.1 Suggested Solutions

In the sections that follow, various solutions are proposed for the problem areas identified on the basis of the comparative analysis.

4.1.1 Problems of non-centered Cameras

Batman Arkahm Asylum (Rocksteady Studios, 2009) presents a robust hybrid solution for games where the camera is not centered. As the player explores the world, the camera is offset to the right. In this case the player can control the camera himself. As soon as a fight begins, the camera zooms out and centers itself so that the player can attack the opponent unhindered. 

Fig. 11: Batman Arkham Asylum - Camera Centered in Battle

Furthermore, the camera of Batman Arkahm Asylum (Rocksteady Studios, 2009) centers itself while the main character sprints. If the main character is running comfortably through the game world, the camera is close to the character, so that all objects can be seen exactly. When sprinting, however, it's probably more important to keep the overview so that the player can see where to flee. 

The problem with the third-person camera in Elder Scrolls Online (ZeniMax Online Studios, 2014) caused players to create a fan-created add-on that implements a camera control similar to the one described above (Esoui, 2017). As soon as the main character hits an opponent, the camera automatically centers itself, but without zooming any further out. The fact that players have created such an add-on on their own initiative and with their own means shows that this is an annoying camera problem that is basically easy to fix. 

4.1.2 Concealed Cameras Problem Areas

In the following sections, solutions to the problem of the concealed camera will be suggested.

4.1.2.1 Geometry Dithering

There are many trees and plants in the world of The Witcher 3: Wild Hunt (CD Project RED, 2015). So as not to obscure the camera by the fauna, the game has a process that makes objects close to the camera appear transparent. The view is no longer obscured for the player and the progression of the camera appears smooth. The development studio called Gears for Breakfast also makes use of this process, which they call "Geometry Dithering" (Mashable, 2017). However, this term does not seem to be established, as no other materials can be found.

The following illustration from the video game The Witcher 3: Wild Hunt (CD Project RED, 2015) shows how the bushes become transparent and how the main character can be seen through them.

 Fig. 12: Witcher 3 - Geometry Dithering

4.1.2.2 Silhouette

In Super Mario Sunshine (Nintendo EAD, 2002), the main character can often be obscured by walls or objects. This is intended because there are objects or switches behind the walls. A dark silhouette shines through the wall so that the player doesn't lose sight of the main character. 

This is not a camera problem as such, as the main character and not the camera is cover-ed. At best, this could also be used as a solution for a hidden camera.

Fig. 13: Super Mario Sunshine - Silhouette

4.1.2.3 "Cut-Out" View

Video game cameras often have a "Sphere Collider" which causes the camera to bounce off objects. This can lead to numerous problems, as explained in chapter 3.1.3. 

The video game For Honor (Ubisoft Montreal, 2017) contains a suggested solution to this problem. There the camera can fly through objects and walls. In order for the player to still see the game world, a special view is used (called "Cut-Out View" for this article). At the edge of the screen it is indicated that you are actually inside the object with the camera. 

Fig. 14: Cut-Out View

Thanks to this solution, a main character can walk through narrow alleys with a streamlined camera without the camera bouncing off anywhere. The player can also always see what's happening in front of his main character in the game. 

4.1.3 «Camera Whiskers»

A Hat In Time was an Indie Video Game under development at Gears for Breakfast. Its developers have published their solutions to the existing camera problems. Based on an interview with Mashable they show that they use a process called "Camera Whiskers" (Mashable, 2017). This allows the camera to detect if the main character is near an object. The process then defines how the camera can bypass this object so that the player's view is not affected.

4.1.4 The Problem of Nausea

According to video game developer Tsurumi Roppyaku, a stabilized camera is the best solution to prevent nausea (Isbister & Schaffer, 2011, p. 172). The player gets sick if the camera moves too precisely with the movements of the main character. For example, a camera can be used as a solution that follows the main character, but only moves when the main character has exceeded a certain height with its movement in the respective viewport (Isbister & Schaffer, 2011, p. 172). If, for example, a player jumps, the camera moves with a delay, i.e. above a certain height of the jump. 

Technically, this can be achieved with the help of a camera boom (Isbister & Schaffer, 2011, p. 172). In the Unreal Engine 4 software, this looks as follows:

Fig. 15: Unreal Camera Boom

The result is a camera that is connected to the main character via camera boom. It doesn't follow the main character too precisely, which can be regulated by programming. The camera boom also exists in reality. The camera boom is controlled by a human person, as the following picture shows:

Fig. 16: Real Camera Boom

The camera boom in Unreal Engine 4 actually works in the same way. In reality, the person controls the camera with a boom. In video games, on the other hand, the player controls the camera using an analog stick, which is then connected to the main character via the camera boom. 

In this way it is possible to implement a camera that follows the movements of the main character without making the player sick. This is not a fixed camera, but a hybrid camera behavior. The player has the possibility to control the camera that follows him to a certain degree. The movements of the camera, however, are regulated by appropriate programming. 

Another point that must be taken into account when developing a video game camera is its speed (Isbister & Schaffer, 2011, p. 171f.). A camera movement that is too fast or too slow can also lead to nausea. The difficulty in setting the speed is to find the perfect balance. Presumably this can only be achieved by frequent quality testing and constant trial and error.

4.1.5 Fixed Camera

Super Mario 3D World (Nintendo EAD Tokyo, 2013) is a third-person game that uses a fixed camera of the Automated type. Although the earlier games of the 3D Mario series used a free-moving camera, this addition to the series has a fixed camera. 

Such a fixed camera gives the player a complete overview of the game world. The advantage is that casual players do not have to take the trouble to control the camera manually. Furthermore, the player can better estimate the distance between two plat-forms he has to jump over thanks to the better overview. 

The fixed camera is often underestimated, although it offers numerous advantages (GDC Vault, 2014). It results in a more stable image, preventing many of the problems mentioned in Chapter 3. While in games with freely moving cameras objects often block the camera (unintentionally), the fixed camera always offers a clear view of the game world. Due to its fixed position, which is determined by the game developer, it does not bounce off other objects. Thanks to the stable camera, a player doesn't get nauseous while playing.

God of War (SCE Santa Monica Studio, 2005) is another example of a fixed camera. If a fixed camera is used, the developers have the option of designing the view in such a way that, as with God of War, a cinematic feeling is conveyed (Hellblade, 2016). However, this means that every part of the camera's environment must be adapted, which may cost many additional development hours. In addition, numerous quality tests must be perfor-med.

Another challenge is to make the transitions between several cameras seamless and fluid (Hellblade, 2016). Example: The main character is in a building. A camera is used to represent the inside of the building. The main character then climbs out through a window. So the camera has to be changed from the one inside the building to the camera outside the building. In God of War this is represented by a "hard cut", which is often used in the film industry (Hockrow, 2014, p.120). This changes from one scene to the next. 

However, a fixed camera does not allow a player to freely explore the game world. Since in some games with such a camera (see Super Mario 3D-World) the zooming is not provided, the player cannot view the objects from close up. There is thus a greater distance between the player and the main character, which may also have an effect on the storytelling aspect.

4.1.6 Job Position of the Person responsible for the Camera

The question arises as to who is responsible for the camera system as part of a video game project. Normally many departments are involved, e.g. programmers, artists and UX de-signers.

The development studio thatgamecompany has an employee named John Nesky who carries the title "feel engineer" (develop, 2014). He was responsible for the camera system in Journey (Thatgamecompany, 2012). This is a position created specifically for the development of the camera system. 

One of the advantages of such a position is that one person can keep track of the development of the camera system. In addition, they constantly test the camera of the game and pass on feedback to the people involved in the development. On the other hand, this professional field increases the number of people in the development team, which massively increases costs. It can also lead to increased coordination effort.

Chapter 5: Conclusion

The comparative analysis has shown that even in the latest 3D video games there are numerous camera problems. This is irritating because there are more and more technical possibilities. The new video games inspire optically by their ever better graphics, however, the camera systems and camera views seem to be given too little attention. In addition, many older video games show how camera problems can be solved or limited. These problem-solving approaches could also be implemented in new video games.

The cameras in third-person games are often located at shoulder height of the main character, slightly offset to the left or right. This causes problems when focusing on a specific target. As a solution, the camera can be temporarily centered when approaching a door or during a fight, for example. The camera can then return to its previous position.

It has also been shown that the camera is often covered by objects.  One solution for this is "Geometry Dithering", which makes objects appear transparent. Otherwise, the camera can also function in such a way that it can fly through objects, whereby a special "cut-out view" can be used to indicate that the camera is currently in an object. In certain games, there are also solutions that allow the camera to dodge the objects ("Camera Whiskers").

In order to prevent the player from getting sick, a camera that is as stable as possible should be used. On the one hand, a fixed camera can be considered. Otherwise, a camera that moves with the main character with a delay is also conceivable. A so-called "Camera Boom" can also be used. 

Finally, it should be checked whether a separate position should be created in the development team for a camera manager who keeps track of the development of the camera system and has a coordinating effect. 

List of sources

a) Bibliography

b) Internet sources

c) Video Game Directory

Originally published on https://www.neogamedesign.com/.