After seeing Avatar and playing with my own Parrot AR.Drone, I decided to make a video that notionally demonstrates why 3D and video were going to be merging in the near future. This video isn't low budget, it's “no budget” (as Ryan Connoly likes to say). The video is here: http://www.youtube.com/watch?v=nlgWnQ4yTYk

The video is pretty simple. A car runs through a stop sign. A police car stops the suspect car, and decides to surveil it using a drone. The drone sends a video feed back to the officers. The “notional” part is simply that, in the future, that video feed can be converted to a 3D representation. Photosynth does this today for stills, creating a 3D point cloud from 2D photos. I had been working with Microsoft's Pivot and Photosynth, and I was impressed in general with their implementation of both products' underlying deep zoom technology. You can see the Photosynth representation of the crime scene at this URL, which is also displayed at 1:15 of the video:

http://photosynth.net/view.aspx?cid=f65bd963-0100-42c0-83b1-c15589f24154

We had a lot of fun filming and producing the video. The only “point” to the video is that we were able to demonstrate this capability, notionally, at an incredibly low price point. We used our own cars, my iPad (first version), my Parrot AR.Drone, my Panasonic Lumix DMC-LX5 and all free software tools. We think the future isn't too far away. We'd love to get your feedback on what you think the future looks like for 3D and video, and whether you think deep zoom-like technology will play a part. Please share your thoughts.