One of the more impressive videos I’ve seen, it combines video projection, a moving camera, various people, props (even fishing lines) plus some surprising technology for this one-take video.
I found the whole video immersive, particularly during the Tron/X-men/Matrix-esque segway. Not only was it a visually interesting video, I was wondering how the projectors managed to correctly display what we saw since the video camera kept moving. For example, if we froze what the video projectors were shooting at the set at an angle, then moved the camera itself around, the surfaces on the set would look strange/offset. I kept thinking it was doable if the video camera moved at a fixed path, with the computer/projector showing corrected images for that particular path.
If you got a little lost, this problem/approach is succinctly showcased in the movie Mission Impossible: Ghost Protocol with the “Kremlin Corridor” gadget. When the protagonists want to travel down the corridor to the safe, they put up a fake wall that projects an image for the security guard at the end of the hall to see. It couldn’t be a still image, though. The fake image of the empty corridor needs to be constantly updated/adjusted to the security guard’s eyes so the corridor looks correctly proportioned at every angle the security guard would see from the actual corridor.
With that, here’s another video utilizing the same technology: hope you enjoy, and see if you can’t figure it out:
How they did do it is actually both simple and brilliant, particularly for this ad. Instead of a fixed camera angle, the cameraman is given license to move the camera around. The projectors show corrected views to the camera, and the 3d projections are calculated and display to (in real time) by a computer and data on where/what/how the camera is seeing. How does the computer know what the camera is looking at and at what angle? A couple of sensitive, accurate Playstation Moves attached to the camera rigs, of course