« Picture Imperfect – Chasing Third Person Cinema | Home | We’ll be Right Back! »
Picture Imperfect – Floating Head Has Gun, Will Game
By Corvus | September 10, 2008
It is difficult to talk about first person camera use in games without spending an inordinate amount of time talking about the first person shooter genre, which I find to be depressing pointless lackluster not to my taste. Fortunately, there’s not much variance in the implementation of the first person camera, so we can quickly cover a bit of history, summarize the common usage, and then move on to various attempts designers have made to improve upon it.
Early first person perspective games typically took place in flat mazes with uniformly wide corridors. Upon entering the maze, the player was presented with a static image of the environment that may, or may not, have an animated overlay depicting characters you could interact with. If you’ve ever seen the SSI Gold Box AD&D games, you’ll know exactly what I’m talking about, although this style of game dates back to the seventies with Maze War (pictured above).
Ultima Underworld: The Stygian Abyss (UW) was the first game to significantly advance this use of the first person camera by moving fully into the third dimension. They got so much right and several things wrong. If we are to focus exclusively on the camera, there are two obvious things they got right. One, the camera simulates walking by bobbing slightly while moving. Two, you can change the camera’s pitch, rotating it up and down, but not so far down as to notice that you have no body. Immediately evident are also two things they got wrong. One, and this is due in part due to hardware limitations, the camera view is not full screen, but wrapped in a rather sizable UI, which reduces the sense of immersion. Two, and this is due in part to the developer’s history creating flight simulators, camera translation (horizontal movement) is driven by the mouse, not the keyboard, which caused frustration among portions of its audience.
John Carmack saw a UW tech demo in action and thought to himself, “I can make that work faster.” And with Wolfenstein 3D (W3), he did. Faster, full screen, more intuitive controls. What he sacrificed was camera pitch, camera bob, true 3D architecture, non-uniform lighting model, and rich environmental interaction. It was a rather expensive tradeoff, but the gaming public didn’t seem to mind. It was a PC gaming experience that was fast, ran on minimal hardware, and let you shoot Nazis. The true clincher to the game’s success is that they released a free shareware version of the game that consisted of the first 10 levels, and everybody was able to get a taste. W3 was supplanted the following year by Doom, which one year later gave way to Doom II, and two years after that saw the release of Quake.
While the subsequent game eventually added a few of the missing elements back into the mix–true 3D architecture, more robust lighting model, full camera control, and, in some cases, camera bob–it did not add in the one key element that UW used to make the player feel part of the world–rich environmental interaction. But that’s an entirely different complaint that is in no way related to the use of camera, so I’ll move on…
W3 established a fixture of the FPS genre that has, in my opinion, drastically hindered our perceptions of what a first person camera means–the ever-present gun floating at the bottom of the screen. Tomorrow’s post is going to discuss this gun and it’s faithful companion the targeting reticule in more depth, but for now I’d like to simply point out that when a gun is a permanent part of the camera view–it seriously informs our approach to dealing with the world.
The biggest challenge facing the use of the first person camera is how to deal with the concept of physical presence. Our primary cultural influence, the FPS, typically portrays its protagonists as a floating head with a gun. Not only can you often pitch the camera straight down at the floor (you try that without leaning forward), but doing so most often reveals, at best, a small round shadow. This reduces your entire physical presence to a weapon and a shadow.
Games ranging from Jurassic Park: Trespasser (JP:T) to Thief: Deadly Shadows (T:DS) to the new Alone in the Dark (AitD) game have made attempts to compensate for the disembodied nature of the first person camera. JP:T, otherwise a physics experiment gone horribly awry, used cleavage to mask the lack of physical presence. T:DS was actually designed with a third person camera in mind, but included a first person camera for the fans of the series. When implementing the first person camera, they included the avatar’s model, so looking downward revealed legs and torso. AitD features an interesting inventory system that’s represented by the protagonist’s pocketed coat lining and utility belt. I’m not sure whether the protagonist’s torso is ever in the first person camera view otherwise, and I know the game includes a third person camera as well.
What no first person camera has ever really done (as far as I’m aware) is to successfully replicate what it’s like to see the world from a first person perspective. Sitting at my desk as I type this, I can see the rims of my glasses, the bushier bits of my moustache, my hand on the keyboard, and bits of my legs in my peripheral vision. When I shift my shoulders to stretch my back muscles, I can briefly see my shoulders. No game recreates this experience and I’m not sure it should. But all these bits of visual awareness of self are completely lacking from the first person camera.
It isn’t just awareness of my own body that informs my sense of physical presence. All of the objects on my desk are visible within my peripheral vision. Far to my right (T.S.) Eliot, Poodle, lies on the living room floor expressing his concern over the impending move by obsessively licking his hind leg. To my left, I can see the white switch plate against the deep red wall of the hallway that connects my temporary office space to the back of the apartment. These are distant, secondary perceptions that inform my sense of presence in my environment. If I were to strap a camera to my face so that I could only see through its lens, the sense of claustrophobia would be overwhelming.
Many games, most notably Half Life (HL), Metroid Prime (MP), and Halo have narratively justified this narrowing of perception by placing full face helmets on the protagonist. MP, in particular, adds to your sense of character, if not physical presence, by showing you brief reflections of Samus’s face when explosions rip across the screen. Water droplets also trickle down the helmet’s visor as you emerge from the water. Full face helmets also rationalize the use of UI elements such as damage indicators, health levels, and ammo quantities–all UI elements that ultimately try to compensate for the lack of physical presence and awareness.
Now, I recognize that my bias is for third person cameras and that many others actually feel more deeply immersed in games when viewed through a first person camera. But until such time as our technology allows us large enough resolutions and affordable immersion, the first person camera will be a limiting factor that game designers must try to compensate for–either through scaling back the environmental interaction, adding UI elements, or visual tricks to convince us we inhabit a world in which we have no body.
Tomorrow’s post will talk more in depth about UI elements and their relation to the camera, but for now–let me have it in the comments!
Tagged:camera, first person, video games. | 18 Comments »
September 10th, 2008 at 10:09 am
“Sitting at my desk as I type this, I can see the rims of my glasses, the bushier bits of my moustache, my hand on the keyboard, and bits of my legs in my peripheral vision. When I shift my shoulders to stretch my back muscles, I can briefly see my shoulders. No game recreates this experience and I’m not sure it should.”
I definitely don’t think games should replicate some of these elements unless these elements are removed from the player’s reality via goggles or something. It would be disconcerting to see my own nose and eyebrows and then look on the screen and see the character’s. I don’t really have a problem with the first person perspective lacking those anchor points because I already have them.
One thing I cannot thank you for is mentioning Trespasser and bringing back the painful memories of the hours I spent playing it.
September 10th, 2008 at 10:30 am
I don’t believe that FPS will be able to ever replicate the feeling of “being a character”. At least, not until they’re able to dissect the gun control from the head control. I don’t know about you, but if I walk around holding a flash light, I might be pointing it in one direction but looking in another. In addition, when your eyes dart from one place to another, picking out details and moving on, your head may never have moved. While you can dart around various locations on the screen, this doesn’t, as you’ve pointed out, take advantage of the human peripheral vision.
Regarding the details of your own face and body which you pick up, I don’t believe it’s important to mimic viewing from a game character’s eyes. You can’t replicate blinking, for one, and there’s no need. In addition, placing details such as the tip of your nose, fails to take into account the various focal points. Would you render the nose as though you’re looking past it, thus in two misty forms that blur out, or as a solid pointy nose, which is impossible to see in life unless you’re completely cross-eyed, and thus suffering a migraine.
Haha, was just thinking, the twin joystick controls for Mech Warrior probably better mimic the functions of the human head, body and eyes than any FPS to date. One stick for body, one stick for head.
September 10th, 2008 at 10:38 am
I’m one of those people who can feel immersed in either, and prefers the ability to switch.
For instance, in Oblivion, I found myself wanting to stay in first person with melee oriented characters, and wantings to stay in third person for ranged attackers and spellcasters. That damn reticle disappears and is just useless when in third person, however–which is why I’m happy to see its use in Fallout 3 seems to actually be worthwhile.
One instance of Bioshock that somewhat bothered me is in the very beginning one is swimming toward rapture, having just survived a plane crash. For some reason, water droplets are on my camera. While it looked awesome, in hindsight, I wonder how these water droplets appeared. I wasn’t wearing a helmet, and if water droplets appear as such on the eyes, I’m actually concerned for the anatomy of my protagonist. Later on we learn he’s a bit different, but even that seems a stretch.
This also made me wonder about playing Deus Ex, as I am often somewhat taken aback when I do see JC Denton. Largely because I feel disconnected from his demeanor and choice of words in the cut scenes from my manner of playing him. And, as Navarre illustrates, my choice of stealth on a character wearing a trenchcoat is somewhat odd. However, when playing the game, unless I run into a mirror, I never even think about that physically damning piece of clothing.
September 10th, 2008 at 11:25 am
Dear Corvus,
Please LRN2 FPS.
You know, my own post ended up touching on Daniel Purvis’ comments quite closely. I even thought of the Mech Warrior style torso-twisting controls. Daniel, you’re so right that they are still far and above anything in Halo, Half-Life, et al…
If I’d say anything about your post it’s that, in being so historically comprehensive, I didn’t get a sense of the *point* for the First Person Perspective. Each to his own, but I still maintain that a first person perspective is the (so far) best way to depict what we would see if we are trying to place ourselves in the game world.
It’s not perfect – my post is actually more about what’s *wrong* with it than right! – but if we are trying to simulate a world (as I prefer to think of videogames, an idea you are well within your rights to disagree with) then you can’t go past the immediacy of a first person perspective.
Dang, it’s gotten late again and I’ve rambled on a bit. Sorry.
September 10th, 2008 at 11:41 am
Hey now! I can FPS with the worst of them. I’ve played and enjoyed all the Half Life games, owned and played the endless iterations if id Software shooters, as well as most of the Unreal games. It’s just that when discussing their use of camera… meh. Not a lot to say.
Friday’s post is going to do a contrast and compare between 1st and 3rd, touching more on the why and how of each camera’s immersive approach.
September 10th, 2008 at 1:57 pm
Denis touches on a good point, I think, which is that first-person games have typically been much more rigid in their use of the camera. I believe I’ve written before about how that affected the innovations of the FPS genre, like Half-Life’s scripted sequences, since they’re reluctant to pull out for an actual cinematic.
As such, because FPS awareness is so focused by the camera view, I’d say that the innovations aren’t taking place in the camera, but in methods of directing player attention. Ultimately, who’s to say that those skills aren’t more important to the narrative growth of the medium?
(not that they’re intrinsically tied to first-person cameras. But the camera style accelerates their development through necessity.)
You’ve raised the point, and you’re not wrong, that many games are strangely unresponsive to player behavior when they use this style of interaction–why doesn’t anyone notice Gordon wandering around the room while they’re talking? Which I guess is fair enough. But I can’t say that I ever wanted to do that while the cinematic was taking place, necessarily, so I always thought it was kind of a moot point. Development of good story and psychological techniques helps to mitigate the flaws for as long as we lack the technical ability to create more dynamic means of storytelling.
September 10th, 2008 at 2:07 pm
Just wanted to quickly say that, while I won’t have time for lengthy replies today, I’m loving the points you’re brining up and will be responding one way or another before this is all over!
September 10th, 2008 at 2:09 pm
Also, if you haven’t already, be sure to click on Ben’s name and read both of his excellent camera posts.
September 10th, 2008 at 2:16 pm
I was just reading in EGM the other day regarding Mirror’s Edge. The designers talked about the mistake of most older FPS titles about placing the camera on the chest instead of the head. The other detail was making the field of view 90 degrees, now I’m excited again for Mirror’s Edge.
My biggest pet peeve when dealing with first person cameras is never being able to see the character’s body when I look straight down, it really pulls me out of the world when I’m just a floating pair of arms.
September 10th, 2008 at 3:00 pm
Minor nit-picks first; it’s Thief: Deadly Shadows, and saying it was designed for a third person camera might not be entirely accurate, as that was something only ever mentioned during the last year of development, where as the ability to see Garrett’s body was first mentioned much earlier.
I’m probably as biased as Corvus, but in the other direction, however one FPS that I feel needs to be mentioned because it made very good use of its camera was Call of Cthulhu: Dark Corners of the Earth. Basically it used the limited information available through a first person perspective to heighten the tension of the game. It’s also full of good examples of physical presence.
The first example I can think of involve the player being chased by a Shoggoth, you have to keep moving down a hallway, opening doors and bolting them shut behind you or you’ll get caught. Because you can’t afford to slow down and look behind you, it’s likely you’ll never actually see the Shoggoth itself, and somehow that makes it even scarier. It’s right there behind you but looking back will kill you. Even though I’m the type of person who would sacrifice my life to just see how they’d animated the Shoggoth I was so caught up in that moment that I was unable, unwilling to turn around. If it had not been for the strict limitations of the first person camera I doubt that scene would have had the impact it did.
Another scene involves your character receiving electro shock treatment, as torture. While it happens you are strapped to a bed, the camera shakes and blurs, and you can hear yourself screaming. The government official interrogating you walks around the room, moving behind you head and out of your vision while still talking, and the fear of what he will do next is enhanced every time he leaves you sight.
A much later example sees you onboard a ship attacked by Dagon, the ship is rocking and pitching in the waves, and Dagon is clawing at it. When a particularly large wave hits, or Dagon gets purchase you have to reach out and grab a handrail to prevent yourself being throw overboard, and you physically see your arm holding onto the rail each time.
Also, like the upcoming Far Cry 2, wounds require that you stop and manually treat them. When you do so you can see your character applying the bandages.
In fact there are simply dozens of examples in that game where the first person perspective has been used to heighten tension and physical connection with the world. I’ll admit my bias but I find it very unlikely that the use of a third person or cinematic camera would have done anything but diminish the evocative power of that game.
September 10th, 2008 at 3:17 pm
Hmm, I remember playing an early FPS game called <i>Catacombs 3D</i>, which may have predated <i>Wolfenstein 3D</i>. (I’ll have to check the listings on Mobygames or something to figure out the exact chronology.) Just to throw one more thing into the mix, there are also the vector-graphics games like <i>Battlezone</i>.
Back on the topic of cameras and views, it’s interesting to note the rise of the third-person adventure as a separate genre. I suppose those games are really the descendants of the 2D platform games rather than the first-person shooters, but then there are hybrids like <i>Jedi Knight</i> and <i>Halo</i> which utilize both views for different aspects of the games.
About presence of self, I noticed a curiosity recently when (re)playing some older shooters, which is the presence or absence of the player’s weapon. This even applies to space shooters like <i>Freespace 2</i>, which features an abstracted HUD. Contrast this with a game like <i>Wing Commander</i> or <i>X-Wing</i>, where the cockpit is visible (but can be turned off) and presents its own challenges, like a partially obstructed view.
In some games, the obstructed view is actually a very real tradeoff used to balance out powerful weapons in some games. That is, you can take the heavy-duty gun, but in doing so you sacrifice some of your visual field. Such obstructions can even affect gameplay, in situations where the reticle is on-target, but where the weapon itself if obstructed by the terrain. That is, peeking left around a corner is fine, but peeking right around a corner can result in a rocket launched into the wall.
Presence of self also manifests as footprints in some games, made all the more curious when you can’t see your feet. Speaking of feet, first-person jumping puzzles are such a pain to deal with for this very reason.
September 10th, 2008 at 3:38 pm
What about Mirror’s Edge? It’s going to be an interesting experiment to say the least.
I remember reading about the first reactions to the camera and how people were turned off right away. The developers then realized that they were tying the camera to the head instead of the eyes of the character. That simple change made doing a roll on the ground a less disorienting experience. Plus, they added a point in the middle of the screen so the player can avoid nausea by looking at it (a trick they learned from ballet: when spinning, the ballerinas always look at the same point as long as they can)
Oh, and the character’s body enters the field of view quite regularly, which seems a necessary but nice addittion.
September 10th, 2008 at 9:38 pm
@Justin
I haven’t played Cathulu, however, when other games have implemented restricted vision and tried to create a sense of violence and tension through camera effects, I actually become disconnected with the character.
I’m not feeling the pain. I’m not screaming myself. It’s not affecting ME. It’s happening to a character I’m playing. In other words, while I’ll appreciate the effort the designers have gone to in order to place the player in the character’s position, their efforts are more likely to yield from me, “oh, that’s a nice effect.” Rather than, “HOLY SHIT THEY’RE TORTURING ME!!!!!”
September 11th, 2008 at 1:25 am
Thanks for the plug, Corvus. =)
Josh – I am totally on the same page with regards to wanting to see my body in first person perspective. I think part of the reason I love Halo 2 so much is because it does this in a commendably simple way – it’s not overblown and annoying (it so easily could be), but its there so that when you are angling down you see your legs and feet.
Kimari – I *really* want to be excited for Mirrors Edge but I just can’t see myself getting the same visceral pleasure out of simply the act of running and jumping as I did with Assassin’s Creed. I was watching a recent trailer and just wishing I could see more of the environment so I could plot my path better. Admittedly, they’ve acknowledged this is an issue and have the red-coloured objects as a guide, but I can’t help but still feel uneasy about the game. I sincerely hope I’m wrong – it looks fantastic.
September 27th, 2008 at 6:40 am
A late comment, but: if I remember it right, in Operation Flashpoint you could see your whole body from the first person view, and in general it did a pretty good job of modelling how a human moves around. Your gun was mostly low down out of view if you were running or walking normally, but up to more of an FPS view if you stopped to aim; and things like crouching, leaning and dropping to the ground and rolling or crawling felt right. You could control where you were looking independently of your movement; I think there was even trackIR support.
My memories are probably mixing it up a little bit with Armed Assault and Hidden & Dangerous.
October 13th, 2008 at 5:32 am
There is a problem inherent with attempting to discuss first person camera use in games without playing alot of first person games in the first place, but you seem to be aware of that already.
From the comments it seems like there are no tactical shooter fans commenting on this blog. Many of the problems you have raised (which are very true) concerning first person shooters i.e player being a gun and arms, not being able to see ones body, having the gun and camera slaved to eachother, are simply not present in the best tactical shooters. Take Operation Flashpoint and Armed Assault for example. In these games the camera is stuck to the player models head and you have two points of rotation, hips and neck. While sitting in your seat angle your body to the left and then look back at the screen, this kind of movement is possible in those games. They also support head tracking for an all around very immersive experience.
October 13th, 2008 at 7:06 am
[...] geschreven, waarvan er twee delen zijn verschenen: één over de bekende ‘chase cam’ en één over de ‘first-person’ camera. Niet al te lang en zeker leuk om [...]
January 23rd, 2009 at 2:15 pm
[...] Elrod has spoken about the limitations of the first person perspective and how it is hardly a realistic representation of the way we see [...]