Virtual Audience Update


**11/12/12 update- If you are interested in the V.A. download, please see this post.**

It’s been a busy day, we shot the ‘audience’ video for the Virtual Audience (see initial post) about a half hour after I finished delivering a keynote speech for the Ohio Learning Network.

After Michele’s initial request, I started thinking about what the possibilities of using Flash and Flash Video could do versus deploying the video on DVD. In doing research for the COM135 Impromptu Widget, I read that Flash can sense movement in front of a web cam by comparing the data from two individual video frames. If there is a difference, there must be movement- or at the very least, changes in light levels. Likewise, Flash can sense changes in incoming audio from a Web cam’s mic. This meant that a web cam could be the eyes and ears of Flash – which could be programmed with some logic, to act like an audience, i.e., see the musician enter, hear him/her play, and tell the audience how to respond when the music stops.

I spent a week or two watching performances and paying particular attention to how audiences act. It usually goes something like this: They wait for the performer to enter the stage. They talk amongst themselves while waiting. When the performer comes on stage, they clap for about eight seconds (unless it’s the Stones or the Beatles or someone like that) then settle into their seats and watch. This is usually where the distractions happen. During music recitals, it’s common for at least one person to clap before the performance has concluded. Anyway, at the conclusion of the performance, everyone claps. A polite audience claps just to be nice. Sometimes a portion of the audience will offer a standing ovation. And sometimes, the whole audience stands.

So, if these audience scenes were shot individually, logic could be programmed into a Flash movie to listen to parts of a performance (which always go in the same order) through the web cam, and load the appropriate video for each audience response. And during the performance, where the distractions take place, a random distraction could load in and play for a random amount of time. All this randomness would prevent a student ‘watching’ the recorded audience for memorizing the pattern of distractions- thus fine tuning the desensitizing nature of the video. In theory that is.

To shoot the Virtual Audience video, Michele managed to coral 125 students into an empty recital hall. I convinced them to sit still for about thirty-five minutes. Long enough to get our distractions and basic video foundations.

Here’s a list of the scenes we shot:

  • talking amongst themselves waiting for the performer
  • welcome clapping- to welcome the performer to the stage
  • sitting and being polite/still (this was difficult- no thanks to Starbucks or Red Bull)
  • gratuitous clapping
  • half ovation clapping
  • full ovation clapping

Then there are the distraction shots. I picked random students out of the audience to perform on cue. There were twelve in all:

  • sneezing
  • coughing
  • yawning
  • falling asleep
  • laughing/giggling
  • crossing legs
  • cell phone ringing
  • entering late
  • leaving early
  • whispering to neighbor
  • clapping early

Stay tuned. There’s more to come as we edit the video scenes and program the logic for the Flash Movie.