Datagram: Live Web Final

Datagram is a live video installation that plays on the meaning of presence in physical and virtual space. It is my Live Web final.


Datagram is an exploration of three themes that I’m interested in: the disembodied user, the virtual self, and the user as voyeur.

The Disembodied User

Much of today’s communications technology focuses on voice, face and hands, and almost wholly ignores users’ bodies. As an aerialist, yoga instructor, avid biker and swimmer, I am interested in how technology can incorporate the body, and counteract our increasingly sedentary culture.

In considering how people engage with their computers, I’ve been inspired by Kyle McDonald’s “People Staring At Computers,” which shows how disembodied the average user becomes when they interact with a screen. His work has made me increasingly conscious of my physical presence in front of my computer. As a result, I often find myself considering how to encourage users to be more present in their bodies when working with screen-based technology.

Working with the Kinect has been a natural fit for my interest in technology and the physical body. It is made to capture the full body, a nice step forward from the webcam, which normally focuses on the face. It also detects movement and depth, which adds layers of complexity to capture data that standard webcams don’t have. Recent work by R. Luke DuBois, Surya Mattu, Wouter Verweirder, and Shawn Van Every to bring Kinect data live online inspired me to develop a piece exploring what it means to be live in virtual and physical space.

The Virtual Self and the User as Voyeur

Studying Live Web this semester has increased my fascination with how people represent themselves and interact when abstracted by the web. Our disembodied self representations created on websites and social media have become so important to our self definition and self esteem, yet they are very abstracted from our actual physical selves.

The abstracted virtual selves sit on the web for other users to delve into without inhibition, creating voyeurs of web users: Each peering into the virtual representations that other users have created. Interestingly, when given the chance, many people hide behind the abstraction that the web provides, acting in ways they would never act if they knew their actual physical identity was involved.

Tricia Wang defends this behavior in her discussions of what she calls “the elastic self.” While I agree with her that safe spaces for exploration of self identity are important, I have reservations about the extent to which self exploration should trump honesty, and respect and compassion for others.


Datagram is a live video installation that plays on the meaning of presence in physical and virtual space. A user approaches a podium where they see a projection of a person inside a box. As they watch the person, they are simultaneously live broadcast into a second box on a podium in a separate place. Users are not immediately—and maybe never—aware of their participation in the experience.

Datagram contemplates digital presence and voyeurism in a world where people interact largely through virtual representations of themselves. The project is in essence a live video chat, yet the removal of the user’s environment, the addition of the Pepper’s ghost effect, and the placement of the user in a box on a podium abstract the user’s image from reality, giving users a sense of how technology abstracts presence.

The Pepper’s ghost effect is used to create a mock physical representation of each user. This image is placed inside a box and behind a clear piece of acrylic—a play on the everyday screens and boxes that house our virtual representations.

Datagram uses live web technology, and therefore can be installed in any two places in the world with a good Internet connection. For example, one podium can be in New York, while the other is in Shanghai.


Datagram is built with peer.js and Kinectron. A Kinect attached to a PC sends a green-screened image of the user to another online user through a peer-to-peer connection. The received image is displayed on an iPad using VNC screen share in a box built to display the image using Pepper’s ghost. The box is placed on a podium.

Although the Kinect is removing the closest user’s figure from the surrounding scene, it does a much better job when it has a black background to start with. Each kiosk has a black curtain hung about two feet behind it to help with this. The Kinects sit about five feet from the black curtain at about knee height.

The Pepper’s ghost boxes are made from cardboard gift and mailing boxes from the Container Store and acrylic from Canal Plastics. They currently sit on podiums built by the amazing and generous Thea Rae.


The installation is currently at an early prototype stage. Here’s a video of it working:

Reflections and Next Steps

The basic functionality of the installation is up and running, but I’d like to develop several aspects of the project before putting it in front of the public.

Live web components:

  • Refactor to run more efficiently.
  • Research better solution to VNC for Pepper’s ghost screen.
  • Consider if image should be further distorted.
  • Consider user experience further: How does this project make sense to a user? Are there prompts? Are there signs? Audio cues?

Physical components:

  • Work on Kinect placement. I think it should be head on, but something will likely need to be built to make that happen—maybe a rig to hang it.
  • Build podiums that allow for charging the iPad and a way to hide or house the computer that is running the Kinect.
  • Situate the black backdrop in a more organic way. Maybe these are rigged from the ceiling?
  • Secure boxes on podiums.
  • Replace acrylic with plexiglass.

Inspiration and Attribution

The works that inspired and added to this project are:

Huge thanks to Shawn Van Every for all of his help working out the code, to Paula Ceballos and Maya Moosh for their amazing support with Pepper’s ghost, and to Danielle Butler for always being willing to play!