My flight out of Chicago was delayed so I mounted the webcam and did an ambient lapse of O'Hare's Terminal 3.

I think some of the most interesting visual features are the consonant and dissonant movement of the crowds, the columns and lines across the ground, the flags, and various reflective surfaces.

Around 1:13 I ran into a cute girl with big headphones and followed her from a distance until 1:38 where I lost her. Somewhere near the end I found her again, she was leaving from gate K2 to who knows where, and my plane was about to leave.

I didn't have a microphone, so I chose a more ambient track that sounded something like the movement looked ("Ahuvati" by Kaki King). In retrospect, I'm most interested in the unintended connections that emerged between the audio and visuals. 1:38 is the only point I synced on purpose, but there a lot of other moments implied.

I'd like to write something that can analyze video for rhythm and harmony the same way we analyze audio. It would be a great tool for transcoding audio to video in the manner of Michel Gondry's "Star Guitar" video or Norman McLaren's "Begone Dull Care". Or, perhaps better, transcoding video into audio?