How Thrilling: Extending The Body

‘How Thrilling’ uses the familiar song and dance of Michael Jackson’s “Thriller” to explore how technology can extend the body into many disparate spaces, through many representations, and for many audiences. Through this lens, the project examines how technology standardizes the body.

The project is composed of four feeds illustrated in the image above.

The performing body is presented through two primary representations in feed 1 and 2 respectively: an abstracted stick-figure-like skeleton that is normalized to a single set of proportions and an unmodified in-situ RGB image feed. The abstraction encourages an unselfconciousness of the performer while also highlighting its irregularity of motion in contrast to the precise repetition of Michael Jackson’s looping skeleton. In juxtaposition, the RGB feed–seen only by an audience in an entirely separate space without the accompanying music–highlights the nonconformity of bodies to any form of standardization.

If the performing body closely matches Michael Jackson’s moves or a set time period expires (whichever happens first), the front projection for the performer switches to reveal a live RGB image feed of the audience watching their RGB image feed. For a brief moment, they can communicate across these displays (basically just like Skype, Facetime, etc.) and the audience realizes they are not watching a recording by a live performance. Then, without warning, the projection for the performer reverts back to the abstracted skeletons.

Two additional feeds provide context within the project. Firstly, a constant silent loop of the original Thriller video excerpt gives visual context to the audience watching the RGB image of the performer. They might recognize the actions of the performer in the Michael Jackson video and vice versa. The last feed visualizes the motion trails of the performing body. Without the skeleton, it draws attention to the impercision of our actions despite attempting repetation.

Desired locations:

  • Feed 1 (Performer + skeleton projection + audio of Thriller song): first floor lobby
  • Feed 2 (RGB image of performer): somewhere on the fourth floor, not too close to the elevators
  • Feed 3 (Original Thriller video loop): somewhere closer to the elevators
  • Feed 4 (Action trails of performer): somewhere on the fourth floor, proximate but not adjacent to the other screens

Bodies in Space

Central Questioning

Why move the body? To control something. Thus, it is not a question of how a body should move but rather why should it move? If the answer is to contol something, what does the “controlling body” control? Either an object or environment.

Central to this question of control is a question of likeness. This is twofold: how closely does the controlled body approximate the controlling body and how closely does the physical extent of the controlling body approximate the corresponding simulated extend? Is one thing imitating another or are the markedly different? (Approximation refers to both behaviour and likeness).

The Body That Controls, Controlling the Body

The body controls as it is detected and performs standardized poses or actions. These map to the action or positioning of the controlled object or extent. This presents two realities: the reality constructed through the control of the physical body (“the simulation”) and the reality of the physical body in physical space (“the controlling”). Through engaging the simulation, our imagination places the physical body in the simulated context being controlled but presents a paradox of the controlling body still occupying itself physical context.

The final project proposes highlighting this paradox by displacing (and disassociating) the controlling and controlled “bodies” across different spaces. Four scenes are presented individually in four distinct and isolated spaces. In this, the performance of the controlling body is seen separately from the performance of the controlled body. When presented in isolation, the controlling body (the primary performance) is seen as both the means and ends, relying on the audience to identify the connection between.

1) The controlling body juxtaposed against the controlled body. There is no audience within this space. 2) A feed of the controlling body performing. 3) A controlled body that is unknown to the controlling body. 4) The outcome, or the controlled body.

 

Forming the Simulated Body: Skeletons and Texture

Skeleton

Skeleton code (P5)
A new skeleton is created by drawing lines between joints that are otherwise unconnected in our physical reality. The “controlling body” toggles these connections on and off by touching particular joints together. As the “controlled body” on screen does not reflect our physical/bodily constrains, it’s interesting to watch the movement of an individual without seeing the digital corresponding counterpart. In watching different individuals, they exhibited a tendency to create recognizable structures (a kite, a heart, a diamond). Additionally, whether the connections which constitute the “controlled body” on screen appeared symmetrical, the behaviour of the “controlling body” was frequency symmetrical: moving both arms simultaneously to touch the head or turning in both knees to connect with each other. Once the individual realized the ability to turn connections “on and off”, their movements became quick mechanical and robotic – hands became entities for touching, grouping the fingers into a monolith.

Some further design decisions to explore:

  • Does the speed in touching the joints together affect the type/quality of line? Are they thick when slow and thin when fast?
  • Do the connections take on different textures?
  • Do the lines fade overtime, which in turn causes repeated touching of the same joints? (A dance emerges through repetition?)
  • Can joints connect to objects in physical space?

Texture

Texture code (PDE + webcam)

Each frame, a single slice is extracted from a random position and displayed over an capture taken when the kinect is initialized. In this way, an image accumulated but it only becomes ‘complete’ or consistent by staying still/constant over a long period of time. The image is contanstly forming and slightly out of our control of the bodies captured within it. When the bodies moved quickly, they were less likely to be captured, but even if they were, it is only partially. In a sense, the body becomes made of “time”.

 

Making a New Skeleton

Drawing Between Joints
A new skeleton is created by forming connections between joints that are otherwise unconnected in our physical reality. These new connections are toggled on an off by touching the select joints together. As the simulated skeleton on screen does not reflect our physical/bodily constrains, it’s interesting to watch the movement of an individual without seeing the digital corresponding counterpart.

Some further design decisions to explore:

  • Does the speed in touching the joints together affect the type/quality of line? Are they thick when slow and thin when fast?
  • Do the lines fade overtime, which in turn causes repeated touching of the same joints? (A dance emerges through repetition?)
  • Can joints connect to objects in physical space?

Link to code