Tools: Max, Kinect, Processing, Skeleton Detection
The subject is to deliver the idea of the time, constant time, which does not exist in out three dimension world. The general idea is to capture the some past moments, and people in front of the camera can interact with theirs or other people’s past moments.
To develop my idea, I made simple sketch, thinking about the interactions and visual forms.
I met some difficulties
The idea uses Kinect depth camera to separate people and background, and only capture blobs in certain depth to make sure it bangs only when people appear. And use the cv.jit.bounds object to get the the bound rectangle coordinates, so it can also get the center of each blob, which is awesomely skip using cv.jit.centroid. When people appear at the beginning of the 60 seconds, the shadow will be recorded, and two shadows would be clear separately at 30 seconds and 52 seconds
Also there are some interesting sub-patches I want to show: create matrix shots and store them in certain file, and retrieve them, then mix them together in a matrix using xfade object. Also how to restrict the bang in certain time to prevent excessive bangs.
Choosing Materials & Effects
The background sound was chosen to be the crowd sound, and when people approach the shadow, two sounds would be triggered, sequently played human voice: “hello”, “I am sorry”, etc. and the landing / back sound. It is trying to make people feel like being in the crowd, and meet the people in the previous moment, and they would say Hi to you, it may be a little creepy, but also inspiring, related to self recognition, self esteem. For example, I may ask myself: Are you the better one than the former second yourself? And sometimes think about myself one day ago, it seems far away.