Your perspective matters: let's talk about cameras in games!


So the long shot thread about God of War has me thinking about cameras in games and what functions they can take in a game!

While a camera is simply a means to give the player vision of what’s going on, the camera plays many roles in communicating a variety of things to the player. A camera can communicate a style of gameplay or where your focus will be, the continent-wide view of a 4X game implies that your focus will be on cities and countries and not so much individuals, for example. Or they can assist certain types of gameplay, you wouldn’t make a fast paced character action game with a first person camera, would you? They can inform the design of the game in positive and negative ways, like God of War’s close up over-the-shoulder camera requiring extra indicators to inform the player of enemies behind them, as well as a quick-turn button.

The camera has so many purposes beyond being a players vision and I thought it would be interesting to discuss some examples.

Discussion Questions:

  • What are some of your favorite implementations of the camera in a game?
  • What are some of your least favorite implementations of the camera in a game?
  • Is there any sort of camera implementation you’d like to see in a game?
  • How about cutscenes? Any good examples of camera work from cutscenes you’d like to talk about?


I do find it interesting that we default to the “camera” analogy when discussing viewpoints in games. I’m curious how much of that is due to Nintendo’s foray into 3D with Super Mario 64, and how they personified the POV as Lakitu with a camera following Mario around. Which, in the game’s logic, makes little sense. I mean, was Lakitu making a documentary or something?

But I digress. I suppose that the use of camera comes from film and TV influences, but it always struck me as a forced analogy, especially when we dispense with it when talking about first person games. In fact, we rarely refer to POV as cameras in puzzle or turn based strategy games, so it really only applies to third person, sports, and racing games. I suppose those are the most filmic of genres, so it makes some sense.

I really don’t have a point here. Talking about cameras in games is weird y’all.


In more recent times, the camera analogy has only gotten stronger. We don’t fake the perception of vision, we fake the camera filming reality.

When you move your eyes to a new point of focus, your brain turns off your vision. When a camera (typically recording all the light for 1/48th of a second before closing for another 1/48th of a second) moves then you get motion blur. Games have worked towards faking the latter, not the former. Tone mapping, HDR, physically based rendering advances: these all work to model a simulation of light hitting sensors and being recorded in the well-understood way that cameras work. The mainstream progression (and this work equally for any non-photorealistic techniques that attempt to capture any other artistic style which is derivative of the camera) is towards making virtual filming possible. Picking the depth of field and field of view so shots can be composed as precisely as in a movie, effectively picking virtual lenses.

An interesting point of conflict comes up from this: the FoV is chosen in a film, and be that an imax screen or on a 14" TV the other side of the room, people watch it as shot (or, at worst, a cropped 16:9 cut of the original). That’s a directorial decision (and different scenes are shot with different lenses). But playing an immersive first person perspective like this can cause motion (simulator) sickness when the area of your vision the screen takes doesn’t match the actual FoV. So some games offer accommodations to tweak this (VR actually needs to get it exactly right for the headset, otherwise simulator sickness is common) so that if you’re playing on a big screen you can widen the “window” to match the wider slice of your vision that the screen takes up. That’s an interesting conflict between a move to always more cinematic presentations and simulating the camera and the things our body actually wants from an interactive look into a virtual space.


Since we’re talking God of War: I like (semi-)fixed cameras. I am used to player-controlled cameras and like everyone else, I’m conditioned with the whole “left stick to move, right stick to look around” mindset, but every now and then I’ll play an older games like Silent Hill 2 or God of War III and be reminded of how great a computer-controlled camera can be when done right.

But if we’re going to do the right-stick camera thing, it needs to be synced up really well with movement. There are too many games where the camera always moves too slowly compared to your character, even after playing around with sensitivity, and it really makes it hard to navigate the world comfortably. I think Super Mario Odyssey does a really good job with this, which is nice to see, because although Super Mario 64 is still my favorite, its camera work was pretty bad.


I thought about what terminology I should use for this topic. Perhaps “point of view in games” is more technically accurate but camera seemed to better communicate what I was thinking about because it’s such a commonly used term.

I would guess the widespread use of “camera” being equivalent to view point does come from the lakitu in Mario 64 at least in somewhat. This is actually a really cool example of point of view in games because the lakitu with a camera was used to help player wrap their heads around the 3D, fully controllable perspective, giving players a foundation for looking around a 3d space. Perhaps heavy handed, but still cool!


Also what is a “Lakitu” :slight_smile: and why would people use an analogy other than a camera for the rendering of (roughly correct) 3D scenes at an interactive framerate?

We can go back to 1980s 3D games and look through exactly how they describe everything but I’m going to bet on the camera analogy being established at that point (although I’m not sure exactly how far back into those early 3D software renderers it was established). I’m certain that we’ve got early flight simulators talking about camera controls because even back then you could do things like change the virtual CRT display in the cockpit to a target camera (and flip on FLIR for at night) so the movement controls when activated for those systems were moving a (virtual) camera. (eg 1987’s F-19 Stealth Fighter - “TrackCam” CRT)

1989’s A-10 Tank Killer refers to the “Floating Camera Views” in the manual and I doubt this is the first time that terminology is used.


To me, a camera is a device that converts incoming reflected light into an image (or series of images) through the use of a chemical or digital medium. As the output of my screen when playing a video game is the result of a long string of data telling pixels to do things, I don’t consider it a camera in the traditional sense. I realize that I’m being super pedantic here, but it’s my petty hill to die on dammit!


So I’ve thought a lot about this too because I feel this is nailed in the Mario Galaxy games. The computer controlled camera reduces the amount of things the player needs to worry about. Considering the planet navigation and all the gravity fun in Mario Galaxy, taking the camera out of the players control is useful in keeping sightlines consistently readable.


One of the few things I liked in Metal Gear Solid V was that in cutscenes the camera was often imitating a camera held by an operator. I really liked that things would happen like models holding doors open for the imaginary cameraman after the player character has already passed through them. I think I enjoyed it because it felt more filmic (if in a very superficial/individually imagined way). Perhaps ironically, I did not care for the narrative told in MGSV despite really liking this device.


Shoutout to Halo (CE) for the first popular implementation of dual-stick first person input curves that were not just absolutely garbāge (combined with level design that was aware of the lower vertical look speed). It made the game playable, an actual alternative to the jump from keyboard to k+m controls for FPS (which, at least for my friends, was around Quake I era - post Duke3D & Doom but before Quake 2) which basically pushed everything forward in terms of immersive input methods. It’s extremely apples and oranges (because k+m involves extreme precision on look and just 9 movement states while dual stick involves precision via straffe movement to compensate for the lack of fidelity for just looking around) but that’s basically the start of my being able to see FPS as a genre collection. Before that then “console FPS” was definitely a completely different genre [in my estimation] that was much closer to pre-mouse FPS of old (five years being a very long time in genre conventions back at the end of the '90s where genres were solidifying and experimenting at a shocking pace).


So I don’t have much to say about the use of the word “camera”, but I will throw in two of my favorite uses of camera, even though they’re relatively simple examples.

First is EA’s Skate.
One of the things that drew me into Skate was, interestingly enough, the camera perspective. The game is in 3rd person, but positioned lower toward the player’s feet, offset to the side, and just a little closer to the player than one would usually find. I don’t know exactly what it is, but just that little shift in perspective compared to the centered, behind-player, slightly raised cameras of other skateboarding games like the Tony Hawk series, I found very visually appealing. I can’t quite remember, but I believe this perspective is actually explained by the game as your friend following you around filming you while you skate.

(For comparison) Tony Hawk Pro Skater 4: 52161-tony-hawk-s-pro-skater-4-windows-screenshot-skate-park

Skate 2: ps3-skate-2-screenshots-1

(P.S. Shout out to the Skate series’ replay editor that’s always recording and allows you to save the last X minutes of gameplay and edit it with a wide variety of camera options. Really cool for a game like this)

Mirror’s Edge is my other pick. I played the demo for this on a whim without knowing anything about the game, and holy crap…A freerunning game entirely in first person? It felt sooooo different at the time, yet sooooo right.


any of y’all play Ico recently? i’ve been playing it for the first time recently, and it has kind of an interesting approach. it frames the action from predetermined angles, but also lets you nudge it around a fair amount with the right stick. i’m not 100% sold on it, because it causes some Sonic Adventure-esque moments where the angle changes, sending you off a ledge, but at its best its pretty compelling.

like, the 3 or 4 minutes after the time this is linked to worked pretty well imo. it feels like a compromise between the creators wanting a shot to look a certain way, while acknowledging that pursuing “filmness” too far could end up being limiting.


As someone who just played through God of War 2005, I say fuck fixed camera angles. There were SO many times that I needed to see more of my surroundings and wasn’t allowed to, because the right stick was for the atrocious dodge roll (which NEVER goes the direction you want).

And if you put bullshit platforming in your game that isn’t a platformer, you’re making an enormous mistake. (Hades fucking sucks, I hated every second of it)

I wish I had a positive example, but the camera is one of those things I don’t notice unless it’s bad.


I found MGSV’s cutscenes to be kinda insufferable for the same reasons. I wouldn’t have minded if it wasn’t so indulgent about circling around characters and military hardware for minutes at a time.

I’m also the guy who thinks that game’s story has a lot of really great stuff to say about western imperialism, when it’s not shoving the camera down Quiet’s top.

I’d nominate MGS3 Subsistence as a great example of the importance of different camera types. Playing the game with either the top-down or the behind-the-back camera exclusively makes for a fundamentally distinct experience. And the ability to swap between them has tactical utility. The game also signalled the end of the series’, and game design in general’s shift away from focusing on spaces to focusing on the body of the player character.

Leaving out every scene with Eva because goddamn they are skeezy, the cutscenes are all directed with an efficiency and verve that none of the other games’ cutscenes have. It’s always looking for a fun way to show us what’s going on and it makes their ridiculous martial art style they came up with seem vaguely useful.


I just thought it was cute and fun to have video game cutscenes emulate a handheld camera. The game play, story, character writing, military aspects, and so on were extremely not for me.


I like Fixed Camera Angles in game when there used well. Its a great way to give the player a sense of space.

Like the Way Resident Evil used it to some times constrict the players. Using it to increase the horror effect of the game.


Attack of the Friday Monsters uses fixed cameras. To keep the player interested and not bored seeing the same perspective of the town. We get different angles of the town which gives you a better sense of it.


I found this video on how Crash Bandicoot worked within the limitations of 3D rather useful for this subject. Among other things it talks about how its uses a mostly fixed perspective to get around a lot of problems 3D platformers have, something Mario wouldn’t get around to until way later:


This is a bit different than the general discussion in this thread, which seems to mainly be about cameras/POVs in 3D space (probably because 2D POVs usually aren’t that significant for these types of questions), but I think 999 is a really interesting example of how a 2D game can use “fixed camera angles” to direct or manipulate a player’s attention. I played through it for the first time this weekend, and it immediately struck me a) how the challenge in its adventure/puzzle sections come heavily from navigating its rooms and spaces, and b) how its developers were able to use usually four or five specific POVs in a given space to direct (or misdirect) a player’s attention to certain parts of those spaces by the angles they choose. I think there’s a bit of a parallel there with @Hunter’s mention of Resident Evil, in that, here, those specific angles produce a correspondingly exacting sense of the game’s spaces.

There’s also a moment in the true ending where the DS’s dual screens start to represent two separate perspectives, one in the past and one in the future, and a final puzzle actually inverts the camera, making you turn the DS upside down to play it. Suddenly, the screen that had represented your character comes to represent a character in the past, and your top screen becomes the puzzle solving screen. It’s hard to describe if you haven’t played it, but, basically, it physically changes perspectives around in a way that mirrors a split happening in the story, and that leads to what I can only describe as a really cool moment.


This is pretty hilarious hill to die on given that this also describes exactly how your screen displays photos as well. Technically, a camera doesn’t display anything.


I’ve been obsessing over this topic since I picked up BATTLETECH so it’s pretty funny to see this thread now.

One of the (many) reasons I picked up the game was when Austin and Rob would mention how you really get a feel for the scale of the mechs (sorry, 'Mechs) especially when a heavy punches a light. But I think the cinematic camera actually ruins that feeling of scale a lot of the time, and not only because it gets confused or stuck. When it spins around wildly, there’s no weight to it or feeling of realism, which just helps you notice all the other places the game falls short in creating that illusion.

People were discussing the “analogy” of cameras, but game cameras in 3D worlds function in basically the same way as normal cameras, they just aren’t limited by the physics of optics. Which is to say, the most important things about them are where they are and what they’re seeing (e.g. FOV). The fact that they have unlimited depth of field or dynamic range doesn’t really matter - there are plenty of ways to get those effects with real cameras if you want to for some reason, and sometimes you’ll use those effects to create a certain feeling in a photo. Likewise, you emulate real camera limitations in games because they mirror things we’re familiar with whicih lets you create a certain feeling, like an illusion of scale. Like @just_benj said about Skate:

One of the things that drew me into Skate was, interestingly enough, the camera perspective. The game is in 3rd person, but positioned lower toward the player’s feet, offset to the side, and just a little closer to the player than one would usually find. I don’t know exactly what it is, but just that little shift in perspective compared to the centered, behind-player, slightly raised cameras of other skateboarding games like the Tony Hawk series, I found very visually appealing. I can’t quite remember, but I believe this perspective is actually explained by the game as your friend following you around filming you while you skate.

The perspective is explained that way because that’s what every 90s skate video looked like. Now that you’re emulating a realistic filming style, you just need to up the contrast and suddenly, BOOM: gritty realism.

In a game like Battletech, you have no skate videos to reference. There aren’t a ton of videos of giant mechs fighting in mainstream culture. So there is a feeling of scale when an Atlas stomps a Locust, but it can look like a big toy hitting a small toy. For example:

What are those big rocks? They’re supposed to be minerals, but are there just boulder-sized mineral deposits on the surface of these planets? They’re the only point of reference we have here for scale, but they’re really only there so you can see what type of position you’re moving to from the top-down camera. The world wasn’t really built for the cinematic camera, so it can’t make things look “cinematic” no matter how hard it tries.

But don’t get me wrong, this game can be absolutely beautiful too (give all games photo mode):

In conclusion: Cameras. They’re great.