Several people have asked if I used any special camera rig to take the pictures. Indeed I did, a very special camera rig, a custom rig in fact. A rig so custom that it's made out of K'Nex:
Yes, that's right, the toy. Here you can see mark II. mark I, used for producing the Engineering Courtyard panorama, was much wider (by two full hub widths) and missing several features present here. However, the basics are the same, the goal is to pivot the camera around the focal point of the lens. The rig accomplishes this by suspending the camera in a framework that provides support for a tilt pivot on either side of the lens. This framework is then placed inside a yoke, that is mounted on a tripod and uses the tripod's existing pan pivot. In order to allow easy tilting of the camera, the framework also includes counter-weights (batteries) out in front of the lens to accommodate the weight of the camera body behind it. The starboard pivot looks like this:
The tan rod is the pivot axle. It's gripped by several parts to provide friction preventing it from sliding out unintentionally (it needs to slide out to remove the camera carrier). Over on the other side it looks like this:v
In both cases the tan clips on the top serve as guides when installing the camera unit. They're largely identical but you'll notice two differences:
There's the view from the side. You can see how far back the camera body is - the pivot point is the tan dot almost dead center in the frame. From above and below the camera carrier looks like this:
The pivot axle is the tan rod on either side of the lens (more visible in the ventral shot). The batteries do a reasonable job of being counterweights, but mk.I required such accurate balancing that just the four hubs clipped onto the dorsal surface were enough to get it perfect. They aren't necessary any more though, as you can see here, the balance issue has been resolved:
On the left you see the #1 most important feature in mk.II: the friction rod. This increases the tilt resistance by a couple orders of magnitude, reducing the balance requirements for the camera unit dramatically. It also provides a nice point for calibration marks to make sure there's adequate vertical overlap between frames. The right picture just shows a few little details of the tripod connection - several thumbscrews have been removed 'cause they just got in the way, and there's some interesting construction down at the bottom.
There's still a few issues with this design however. It doesn't really grab the camera that well (it can yaw a bit inside the cage), the counterweights are overly bulky (There's things far denser than batteries), and there's still way too mucn play in the whole system to make non-registered bracketing reasonable.
Hope you found that all at least a little interesting. If you've got any questions, drop me a line at inioinio@gmail.com.