Thesis
The Project
Living Movies is based in Improvisational Theater. What is Improvisational Theater? It's a performance style that seeks to create characters and situations based solely on words spoken by the actors or the audience. Instead of a script, the action is determined by connections that the actors make with each other.
Central to Improvisational Theater is the concept of offer and acceptance, often signified by the phrase "Yes, and." The idea is that one actor makes an offer: a statement, a physical action or a character, and another act responds in a way that accepts the reality put forth by the first actor and then adds to it. For example, I might make a suggestion by saying "What a beautiful day for a cruise!" My scene partner might accept my offer by saying "Yes but it's my first voyage." And then add to it by saying "I think I'm going to be sick".
Improvisational Theater typically involves the audience in the action in the form of suggestions that are solicited. Usually these suggestions come before the beginning of a scene or a performance. Sometimes the suggestions can come in the middle of the scene, especially in theater games or "short form" Improv.
I based Living Movies on an Improvisational Theater style called the Harold2, in which a series of two person scenes are presented in sequence, with follow up scenes that jump in location or time.
Seinfeld has been cited as the perfect Harold, in which several plotlines intertwine and are eventually resolved.
The action takes place on a video screen as opposed to a stage. The actors are composited with video backgrounds using a blue screen technique.
By using a particular shade of blue, the color can be removed or "keyed out." The background can then be replaced with another video source or graphic. The video of the actors is fed into a video mixer where it is mixed with video coming out of a computer that replaces the blue background with video of various exterior scenes.
There is a computer operator who selects the backgrounds as well as sounds and music.
The audience communicates by means of online chat. Instead of calling out suggestions as in traditional Improv, the audience types in their comments. The chat text is sent into a second video mixer, where it is mixed with the video of the actors and the video backgrounds. The result is then sent out to a video projector. The video of the actors appears against a video background. The chat text appears over the video on the top of the screen.
The software used to select the video backgrounds is a program used by video jockeys or VJs called Grid Pro. This software uses a graphical interface that allows you to select video files and play them by clicking on an icon. The chat application is built on PHP/MySQL and is called ARSC (A Really Simple Chat). The video mixers are basic consumer grade video mixers, the SIMA SFX-9.
This project originally started out as an assignment in Shawn Van Every's "Producing Participatory Media" class. A group of four students worked together to come up with a concept to test.
Although the group seemed to like the idea, there was some concern about how interesting the end result would be. For one thing, if the performers weren't very good the scenes wouldn't come off well. It also wasn't clear how extensive of a database of backgrounds would be necessary. In the end, the group went with a different concept.
When it came time to test the concept for my thesis, I wasn't sure how to proceed. Since the idea required both the participation of a number of actors and technicians and the use of a large amount of computer and electronic equipment, it was very hard to test the concept. Originally, I had thought that I wanted one person to be able to carry and set up all the equipment for the project. I was never able to achieve this, and in practice it took two people to handle everything.
I decided for my first or "Alpha" test to use puppets as a way to reduce the scale of the production. I set up a video camera, shooting the puppets against a blue piece of paper. I then sent video of street scenes out of my laptop and mixed the two sources together.
At this point I didn't have a way of getting the chat text onto the video so I faked this part by using a video of simulated chat text. I tried to create something that might look like the kind of interaction I was aiming at.
For my first Beta, I enlisted the help of some Improv actors who had expressed an interest in the project. The Beta test took place in a classroom at ITP. I prepared three types of videos for the test. The first kind is a backdrop video. These were videos I had shot of street scenes, mostly storefronts. The second kind is an activity scene. An example is a video of myself drinking coffee. The third kind of video is a character scene. I shot several monologues of different characters.
When it came time to do the test, we ran into a few technical problems. Lighting the blue screen was trickier than I thought which resulted in poor video quality. We also ran into some audio problems. By the time I got everything working, it was about an hour and a half longer than I had anticipated. I was only able to get through a few scenes with the actors.
I initiated the first scene by showing a video of myself drinking coffee. This caused Ben to start off a scene about coffee.
After the coffee scene, I changed the backdrop to a church. The actors responded to this change.
Before the start of the second Beta, I ran through the audiovisual setup with Marlon in Technical Services so that things would go a bit more smoothly. I also finally had a solution to the problem of getting the chat text on the video screen. This consisted of: 1) A chat application built on MySQL/PHP. 2) A Perl script that pulls up the last three comments from the chat database. 3) A Meta refresh tag in HTML that reloads the page with the Perl script every three seconds. 4) An Opera browser client running in Kiosk mode so it only shows the page contents - no toolbars, etc. 5) A scan converter that takes the text on the computer display and sends it out to the video mixer.
Not a very elegant solution, but it did work. Actually, when it came time for the second Beta test, I couldn't get the Kiosk mode to work and the toolbar showed up after all.
For the second Beta I had three actors. I also invited people to come watch the performance and make suggestions. I had a computer set up for chat and I actually did have one person chatting online.
I set up the theater so the actors are shot against a blue screen on one end, a projection screen was on the other and the audience was in between the two. On the side I was the computer operator and a classmate, Kesu James, was acting as the online chatter.
The actors took cues from the chat text, for example where the actor uses the phrase "Thursday is the New Saturday" and the chatter replies with "four o'clock is the new midnight."
The actors also cued or "pimped" the computer operator for example where an actor plays air guitar and the operator responds with Green Day's "American Idiot"
Afterwards I got a few comments about the project. The main comment from both audience members and actors was that the text was too hard to read. What had looked good at home staring into a monitor (white text keyed over video) was illegible when projected on a screen. And the bright lights shining on the actors made it even more difficult for them to see. The audience did enjoy the scenes and most commented that they would like to see more.
2 Del Close and Charna Halpern invented the Harold as a combination of Halpern's short theater games and Close's longer form scenes. After they had developed this new theater form, Del asked "What shall we name it?" Charna replied "Let's name it Harold."
Comments or questions? Email to info[AT]henryharvey.com.
Copyright Henry Harvey 2006.