Chungking Espresso

Left 4 Dead, Subverting Horror Genre Conventions

Posted in Game Analysis, Papers, Schoolwork by Simon Ferrari on February 3, 2009

You know how the first big news item about a game (after its announcement, that is) is always a post of beta-build screenshots? In this assignment, I show just how much information about a game one can draw from just a still image.

Assignment: Using the ACM format, analyze a videogame at an unusual level of granularity: the still image.


Left 4 Dead: Subverting Horror Genre Conventions

Left 4 Dead subverts key “survival horror” genre conventions from both film and previous videogames in order to create a first-person cooperative experience more akin to war movies and games. The example image shows how spatial and lighting cues promote visual clarity over the construction of suspense; furthermore, the first-person perspective strengthens personal presence and agency over the horror genre’s typical simulation of helplessness.
L4D, Valve, zombies, videogame, Left 4 Dead, xbox 360
This article deals with the construction of the image in the videogame Left 4 Dead (referred to as L4D). This analysis draws from the fields of visual rhetoric and design in both cinematic and ludic arts. An introductory understanding of 3d modeling and lighting in the reader is assumed.
L4D is an important recent example of the first-person shooter and survival horror genres. Relevant games in these genres include Half Life 2 and the Resident Evil series, respectively. Specifically, my analysis of the construction of this image will show how the narrative and visual trappings of survival horror can be manipulated and applied to create another (the first-person cooperative shooter) experience altogether. This article will focus primarily on how lighting and spatial cues accomplish this goal.
Camera Eye versus Human Eyes
The defining aspect of first-person games is their point-of-view perspective. In film history, it took a half decade before Robert Montgomery directed Lady in the Lake (1947)1 entirely in a first person perspective. In this film, viewers only saw the protagonist’s face when he looked directly into a mirror. This is the same completely first-person experience perfected by Half Life 2. It took significantly less time to develop the first-person perspective in the field of videogames: Maze War (1973) was developed only twenty years after the first graphical game OXO (1952).

The reason we see so few examples of first-person horror games is the importance of directorial control in creating suspense. Alfred Hitchcock stands as the master of this almost perverse cinematic pleasure. Among horror games, Doom 3 stands out here for its low-key lighting and first-person perspective. The Resident Evil series (prior to the 4th, using an over-the-shoulder camera), on the other hand, exploits the third-person camera in order to set up cinematic angles and limit the player’s visibility control. L4D combats this directorial control (despite the presence of the sinister “AI director” that we will explain in our later discussion of gameplay) by allowing players to pivot and turn their field of vision at will.

Unlike film, where the “camera eye” necessarily extends about a foot (the length of the camera) away from the operator, in games the eye can be realistically located in the face of the avatar. It is important to distinguish this “eye” from actual human “eyes:” the image created here is monoscopic (as opposed to the way we see, stereoscopically) [Arnheim, 1974].

The major downside to monoscopic visualization is that it frustrates depth perception – leading to the increased importance in film and games for space and lighting cues [Arnheim, 1974 and Bordwell, 1985].

Space Construction and Cues
We can see how the camera in L4D simulates linear perspective in order to render a rectangular room how a human would actually view it. The strong diagonal view of the room I have used in my screen capture replicates the effect of the first visually dynamic film image: The Arrival of the Train at La Ciotat. A diagonal static image renders actual space far more accurately than looking at the same space straight on [Arnheim, 1974].

L4D constructs an image of both deep focus and depth of field. The degree of focus on the avatar’s hands in the foreground is roughly equal to that of the old man and zombies in the middle ground; however, textures do become less defined the further away we look from the foreground. This progressive decrease in texture aids in the mental construction of depth [Bordwell, 1985]. Beyond the flames and out into the foggy night forest outside, one can make out the trees closest to the player because of their sharpness compared to the deeper forest fading into a haze behind them. This phenomenon is called atmospheric perspective [Bordwell, 1985].

It took Gregg Toland remarkable amounts of image manipulation in order to attain the famous depth of field and deep focus displayed in the famous opening to the childhood sequence in Citizen Kane. In videogames, deep field and focus are only limited by processing power in relation to draw distance. In more recent games this visual clarity over great distances needs to be purposefully distorted in order to create suspense. One example of this purposeful distortion or concealment is the fog of war (players cannot see the current state of a location without a unit nearby) mechanic in most RTS games. L4D instead embraces advances in processing power and draw distance.

Another important spatial cue in this image comes from familiar size of “props” in the mise-en-scene [Bordwell, 1985]. David Bordwell writes that viewers rely on familiar objects on the screen in order to calculate depth. The objects in this room are ideal toward this end because they are what most of us see constantly in daily life: folding tables, boxes of file folders, and waiting room chairs. The dead bodies on the floor also provide key information; we all know the length of the average human body. In videogames these spatial cues provide a tactical advantage as well as aiding in depth – a player relies on this to see how close a zombie is and how quickly it moves.

Horror Lighting Conventions Disrupted
One of the things an experienced player or viewer of survival horror will notice right away on viewing this image is the high key lighting of the scene. Key lighting comes from the avatar flashlights, creating the exaggerated candle power of a spotlight. Ceiling flourescents act as a soft fill and top light here. Despite being reasonably high key, the harshness of the key flashlights does create strong cast shadows on the walls – nobody would mistake this for classic Hollywood three-point lighting [Bordwell & Thompson, 2003]. Another minor lighting element comes from the gun flare; this acts as a subtle backlight creating the luminous contour of our avatar’s hands.

The image does feature some attached shadows on the hands of the player’s avatar and the realistic contours of the other avatars and their clothing; however, L4D avoids the horror genre convention of using expressive attached shadows and underlighting to convey ambiguously sinister motives in other characters [Bordwell & Thompson, 2003]. In the horror genre, other humans frequently pose just as much (if not more) of a threat to the protagonist as the monsters do. L4D’s lighting emphasizes the bond between its heroes – reinforcing for players that they have no need to fear each other.

Focusing on the environment outside the room, one will notice that the light is unrealistically bright and even for a forest at night. Although some of the game takes place in darkened sewers and office buildings feature low-key lighting, much of the action takes place in evenly lit exteriors. Just as in the tradition of using high-key “day for night” filtered lighting in the cinema [Bordwell & Thompson, 2003], the exterior we see through the doorway supports viewer clarity over suspense-building.

Finally, the flames in this image do not realistically distort our field of vision or even produce enough light to disrupt the shadow on the wall in the back corner of the room. One recent game, Alone in the Dark, boasts realistically propagating fire as its primary source of light through much of the game. The fact that L4D ignores this reflects either a difficulty in the programming of such a dynamic light source, or again a choice to not limit viewer clarity with heat distortion and erratically shifting light.

Presence and Agency
Finally it stands to take a look at “presence” in this game world. Although this will become more important when we return to this game as a moving image and interactive experience, the still image does provide a modicum of sensory immersion and mental engagement for the viewer.

Heeter’s examination of VR in the early 90’s distinguishes between first- and third-person VR configurations, and it also parses presence into its personal, social, and environmental modes. Photorealism is not nearly as important as realistic motion and tightly coupled action for creating a sense of presence in games [Heeter, 1992].

The primary method of creating personal presence in first-person VR is the visualization of the user’s hand. In our still from L4D, we see that our hand and gun occupy 1/8 of the entire frame. Its realistic texturing and the lighting mentioned earlier lends believability to the simulation. Janet Murray’s definition of agency – a user being able to make meaningful action with appropriate feedback inside the digital artifact – also informs our reading of the image. Even though our view here is static, the flare from player’s gun shows that we have captured an instance of meaningful action within the image [Murray, 1997].

We can also derive social presence from this still. Horror games typically pit a single human against innumerable monsters. The lack of social presence (zombies are usually assumed to lack reason and empathy) helps create the feeling of solitude and danger. It is obvious that the three humans with guns in this image are our comrades here. The line of sight from all of the avatars toward the old man in the center of the image conveys a common focus and struggle. Despite the presence of a raging wall of flame and intruding zombies, the image feels almost “safe” because of the reassuring social presence of the other human avatars. This social presence is clearly reinforced by the advanced 3d representation of the avatars, explorations into the pitfalls of the “uncanny valley” aside [Atkins, 2003].

Through spatial cues and lighting, Left 4 Dead effectively subverts the genre trappings of survival horror into a cooperative action experience.

The first-person perspective disrupts directorial control over the mobility of the player’s visual field, an essential tool for creating suspense. Deep focus and depth of field allow for an easy survey of the game state. Spatial cues such as texture and the familiar size of common objects enhance the player’s ability to construct coherent space and depth in their minds. Relatively high-key lighting grants players visual clarity, reducing the emphasis on surprise and allowing control of the environment. Finally, the feeling of helplessness typically constructed by horror films and games has been replaced in this game by a sense of agency and relative safety through strong personal and social presence.
Future writings on this game as a moving image and as an interactive experience will serve to support and expand  this conclusion.
[1]Arnheim, R. “Film and Reality,” Art and Visual Perception, (Berkeley: University of California Press, 1974), 8-34.
[2]Atkins, B. More than a Game, (Manchester, UK: Manchester University Press, 2003), 1-26.
[3]Bordwell, D. Narration and the Fiction Film, (Wisconsin, University of Wisconsin Press, 1985), 113-125.
[4]Bordwell, D and Thompson, K. Film Art: An Introduction, 7th edn, (New York: McGraw-Hill Inc., 2003), 191-198.
[5]Heeter, C. “Being There: The subjective experience of presence,” Presence: Teleoperators and Virtual Environments, (Cambridge: MIT Press, Fall, 1992).
[6]Murray, J. Hamlet on the Holodeck, (Cambridge: MIT Press, 1997), 128-132.
[7]Source image acquired from


Leave a Reply

Fill in your details below or click an icon to log in: Logo

You are commenting using your account. Log Out / Change )

Twitter picture

You are commenting using your Twitter account. Log Out / Change )

Facebook photo

You are commenting using your Facebook account. Log Out / Change )

Google+ photo

You are commenting using your Google+ account. Log Out / Change )

Connecting to %s

%d bloggers like this: