I ran across a very interesting tutorial today:
How to create fake miniature environments
Essentially, this tutorial is simulating an effect that can be created by using a tilt & shift lens. (Or an effect that can be created by using a normal lens with a wide aperature to take a photograph of a very small area, or a 50mm lens the size of Tokyo taking a picture of Tokyo, or a telephoto lens several miles long)
I find this tutorial interesting because our brains immediately believe that the thing that is being looked at must be very small. Why? Because many years of photography have trained us to believe that an extremely shallow depth of field can never occur over large, distant planes. This kind of depth of field can only be found in very close macro shots of very tiny objects. We have, in essence, developed a sort of visual literacy when it comes to depth of field. Yes, of course, our eyes experience depth of field, but our eyes are roughly like 18mm lenses, and as such the depth of field is fairly large. The literacy has changed a bit, too, since we've shifted from large format photography (which had much narrower DOFs) to 35mm. Likely, the image below if shown to someone from the 11th century would not appear to be "miniaturized", it would simply appear fuzzy at the top and the bottom, assuming they could comprehend the idea of a photograph at all.
Here's my fun experiment, a shot of San Francisco at sunset from the Coit Tower (click to enlarge):