proloe:

adulthoodisokay:

wait what

Okay but a fun fact about this: There is actually no difference (without stereoscopic depth information provided by us having two eyes) between the image produced by a convex object lit from one direction and a concave object lit from the opposite direction. Literally, they would produce exactly the same photograph. When viewing 2D images like this, our brains just make a best guess, and this is an example of it getting it the wrong way around.

This phenomenon is made clear when doing analysis of photographs for height information. There are usually two solutions when the direction of the light source is not specified, and the correct solution usually has to be manually selected. I’ve had to do this on multiple occasions when collected depth information for video game textures.

I think this is really neat and not something people come across very often (because it’s rare that we come across a photograph devoid of enough context for our brains to guess correctly), and it’s really cool to look at how this works.