The Challenge of Cameras

Cameras in games are difficult to get right because they can serve two very different purposes at once. First, the camera has to serve the gameplay. A wide-angle view gives the player more peripheral vision. A tightly zoomed camera makes precise aiming easier. A two-dimensional view makes it effortless to judge the distance between two characters in a way that can be harder in 3D. A linear game might benefit from an on-rails camera, while an exploratory game usually needs a camera the player can control.

Second, the camera can serve an aesthetic purpose. A very wide shot can make a character feel small and insignificant. A close-up view can make the player feel trapped. A first-person perspective can feel immediate and embodied, while a high viewpoint can make the player feel powerful and detached. Just like in cinema, perspective, distance, and angle can communicate tone, mood, and power dynamics before a single line of dialogue is spoken.

That gives designers a useful but dangerous tool. A camera can subtly push the frame toward a more pleasing image, as in Shadow of the Colossus, where the camera shifts while you ride the horse so the action follows the rule of thirds. The character is not held dead center because the frame looks stronger when the subject sits slightly off to one side. The camera is still playable, but it is also composing the shot.

A camera can also be taken almost entirely out of the player's hands. The classic Resident Evil games used static, pre-rendered backdrops that made the action feel like it was being seen through different security cameras. That approach was partly driven by the limits of the original PlayStation hardware, but it also gave the developers complete artistic control over what the player saw and when they saw it.

The licker introduction in the original Resident Evil 2 shows how powerful that control can be. The first camera placement uses the leading lines of two walls to pull the player's eye toward a window, where something strange moves past. The next angle appears to take the viewpoint of whatever was outside, making Leon look like prey. Then the camera cuts high, making him look small and vulnerable.

The sequence keeps tightening the player's fear by deciding exactly what information is visible. Leon has to run toward the camera and around a blind corner, so the player cannot see the danger ahead. Another shot frames a broken window and a decapitated zombie while obscuring the route forward. Another holds on a bloody puddle, a dripping ceiling tile, and a broken window. The player is invited to move slowly toward the edge of the screen because the camera has turned the edge of the frame into a threat.

That is a very specific kind of horror. The camera does not merely show the room. It withholds the room. It uses composition, blind spots, and angle changes to make the player imagine what is waiting outside the frame. The cost is obvious: the player has less direct control over what they can inspect. The benefit is equally obvious: the designer can make the view itself frightening.

The 2019 remake of Resident Evil 2 made a different tradeoff. It abandoned the fixed viewpoint and moved the camera close behind Leon's head, creating up-close encounters that feel more immediate. That loses some of the staged licker reveal, but it makes a zombie lunge feel more physical. The player is closer to the character, closer to the attacker, and closer to the moment of contact.

The general idea is simple: the closer the camera gets to a character, the easier it is to relate to them. To some extent, that is true. In the 2013 Tomb Raider reboot, the jittery close-up camera does a lot of work in connecting the player to Lara's survival struggle. It frames her body language, stays tight during difficult moments, and keeps the player with her through flooded chambers and vertigo-inducing climbs.

The camera even has a dedicated physics system so it can bounce and bump along with Lara's scrapes. The animation supports that relationship too, often turning Lara's face toward the camera so the player can see fear, strain, and relief. A wider camera could make the action clearer, but it would not create the same connection to Lara's body and expression.

But the closer the camera gets, the less peripheral vision the player has. The character also covers a larger portion of the screen. That can be acceptable in a shooter, because aiming at distant targets is more important than seeing everything directly to the left and right. A behind-the-back camera can offer some benefits of first person while still showing the character, supporting cover systems, traversal, and identification with the hero.

Action games create a sharper problem. The original God of War games used a zoomed-out camera that gave the player a wide view of the battlefield. That fit a combat system built around many enemies, sweeping attacks, and area-of-effect magic. Kratos felt distant, almost like an action figure, but the player could read the fight.

God of War on PlayStation 4 chose the opposite priority. Its camera stayed close behind Kratos, making the world feel tied to his perspective and supporting a more human story. The camera made narrative sense. It also made combat harder to read. Enemies could approach from outside the player's view, and the distance between Kratos and a target could be hard to judge when swinging the axe. The camera worked beautifully for throwing the axe because that action behaves like a third-person shooter, but it struggled more during general melee combat.

This is not only a visibility issue. It changes the emotional experience of play. The older God of War games could make the player feel powerful and predatory because the camera showed the battlefield as something to dominate. The newer close camera could make the same player behave more cautiously, even retreating from the action to avoid being hit from outside the frame.

A close camera can still work in an action game. Hellblade and Dark Souls keep the player fairly close, but those games are built around deliberate movement, reactionary play, and often one major enemy at a time. In a game with many enemies, like Devil May Cry, a wider perspective is usually more useful because the player needs to read everything happening around them.

The God of War team did a great deal of clever work to make combat function with the new perspective. The game uses arrow indicators to show where off-screen enemies are standing. Atreus calls out danger behind the player. Enemies outside the frame become less aggressive so the player can focus on the visible threats. Kratos magnetically snaps toward enemies just out of reach, which helps cover the depth-perception problems created by the camera.

The designers also limited the maximum height of juggled enemies so they would not disappear out of frame, and added a lock-on button so the player did not have to keep fighting the right analogue stick. A GDC Vault talk covers the years of work involved in making that combat fit the new viewpoint. The result is impressive, but the amount of support required reveals the central tension. A camera chosen for intimacy can force the rest of the combat system to compensate.

There may be a simpler answer than forcing one camera to solve every problem: let the camera change. Cameras can move during play. Most third-person shooters zoom in when the player aims, giving them a better view and finer targeting control. Spider-Man pulls the camera far back while web-swinging, making the hero small on screen so the player can see the route ahead. Vanquish holds the camera back during rocket slides to heighten the sense of speed.

God of War does make small adjustments like this. The camera moves in when Kratos throws the axe and hangs back a little when he runs. But other games make more dramatic changes depending on what the player is doing. Batman: Arkham Asylum is a strong example because it understands that a complex 3D game does not need one permanent viewpoint.

When Batman walks down Arkham's corridors, the camera sits close behind his back. It keeps the player attached to the character and the environment. When a fight begins, the camera pulls back dramatically so the player can see the action. It automatically targets and locks on to the enemy Batman is fighting. The viewpoint can swing around in a way that may feel intense, but the goal is clear: frame the fight and keep the most important information on screen.

Bayonetta uses a related trick. In some duels, the camera pulls toward a side-on view, temporarily giving the fight the readability of a 2D fighting game. The camera does not treat every moment as the same kind of play. It adapts to the information the player needs.

Batman uses several other camera modes too. When the player crouches, the game understands that stealth is happening and pulls back to provide more peripheral vision of nearby threats. When Batman takes cover, the game gives a fixed angle that shows around the corner while still framing him well. When he is on a gargoyle, the field of view opens wide enough to see the room. Crawling through a vent switches to first person. Certain Scarecrow sequences become two-dimensional so the player can focus on platforming. Poisoned hallucination scenes tilt into a dutch angle to show that Batman is off balance.

Rocksteady's key insight is that there is no one-size-fits-all approach to cameras in a multi-faceted 3D game. The camera can shift and morph to fit the current need. The player still feels close to Batman because the default view spends so much time behind him, but the camera has the sense to pull back when the player needs more information.

God of War could have used a stronger version of the same idea: stay close to Kratos while exploring, then pull back to a more generous viewpoint when enemies appear. That would not require cutting away from the continuous-shot presentation. The camera could move within the same shot, preserving the cinematic ambition while reducing the combat cost.

The broader lesson is not that close cameras are wrong. The lesson is that camera distance, angle, and control should respond to what the player is trying to do. A camera that is perfect for story scenes may be restrictive in combat. A view that makes traversal exciting may not help precise aiming. A horror camera that hides information may be brilliant in a hallway and unbearable in a brawl.

None of this should discourage developers from trying unusual cameras. If games never changed perspective, they would not have the pulse-pounding action of Resident Evil 4 or the body horror of Resident Evil 7. Great things can happen when designers push against the assumptions around a viewpoint.

Conventional wisdom might say that first-person platforming is a terrible idea, yet Mirror's Edge works thanks in part to a wider field of view and responsive full-body animation. Fighting games are expected to use a side-on perspective, yet ARMS uses long-range attacks and the relative distance of incoming fists to help the player judge space from a different angle.

Tony Hawk's Pro Skater followed the skater, but Skate found a different feel by letting the camera follow the board. Firewatch uses expressive animation to communicate character from a first-person perspective, even though that camera type often hides the player character's personality. These examples work because the camera choice is supported by the mechanics, animation, and feedback around it.

The camera is a powerful design tool. It can support aesthetic and cinematic goals, and it can make a game feel more intimate, frightening, expressive, or fast. But it becomes frustrating when the actual experience of play is harmed. The camera should fit the gameplay, not the other way around.