Clearly I hadn’t! (This confirms my often-told joke about my work in ‘theoretical optics’: “As soon as it can be applied, I change topics.”) The tilt-shift lens is quite cool, though, and it’s now clear how it works.

PD wrote: “To take a picture through a pinhole, you have to either increase the illumination of the scene or use a longer exposure.”

Quite right. Even looking at a single aspect of photography such as depth of field, one realizes how many factors must be taken into account to make a well-crafted image. Aperture size, focal length, lighting, lens type and other factors all contribute, and are not necessarily independent of one another. This complexity explains why most of my photographs end up looking like a Yeti riding the Loch Ness Monster while being illuminated by a UFO.

http://en.wikipedia.org/wiki/Tilt-shift_photography

http://images.google.com/images?q=tilt+shift+lens&ie=UTF-8&oe=utf-8&rls=org.mozilla:en-US:official&client=firefox-a&um=1&sa=X&oi=image_result_group&resnum=1&ct=title

http://en.wikipedia.org/wiki/Perspective_correction_lens

…and a whole lot of light! Something that is implied but not stated in this article is that the more you reduce the diameter of the entrance pupil of the camera, the more less light reaches the focal plane. To take a picture through a pinhole, you have to either increase the illumination of the scene or use a longer exposure.

Part of the reason I included the math was that it forced me to actually do the calculations myself!

Tom wrote: “I think the ones with people are more recognizable as real because it’s difficult to get miniatures with realistic body proportions and poses.”

For me, the most obvious ones are the ones which show too much, e.g. the first one. When you can see for miles into the distance, it’s hard to imagine a miniature of that size being constructed.