We all know – or should know – what a camera obscura is: a pinhole in a screen allows light from outside a darkened chamber to produce an image on an internal surface; consider a situation in which a slide projector projects an image onto a screen with a pinhole in it somewhere, and a second screen behind that. What will the second screen show? Since the pinhole is only exposed to a small portion of the image, it should show only a (dimmer, and (possibly) unfocussed) enlargement of that image segment. Simple raytracing suggests that it’ll show the complete image.

If it does show the complete image, does that imply that every part of the displayed image contains the complete image, like a fractal?

> Simple raytracing suggests that it’ll show the complete image.

No it doesn’t.

Well, that was resolved quickly.

What’s next?

A Camera Obscura does not necessarily need a pinhole, the larger room-sized ones use a lens.

i have a camera obscura. the wood moved a tad. lens is from an old process camera, used in the days when film was the go in printing. top screen is from an old flat screen monitor. and there is a mirror tile at 45 degrees inside.

I worked as a projectionist for a while, and did an experiment similar to the one suggested. The results were complex.

A cinema screen is not a solid layer, but has a regular grid of holes (about 0.2-0.5mm diameter, and about 5mm apart vertically and horizontally), so some of the light from the projector passes through it. There are speaker stacks behind the screen, so it’s not possible to view the whole picture, but it is possible to see reasonably large chunks of it. The parts I could see were the segments of the picture being displayed on the screen at the time, and not the whole picture.

When I chose a single hole and blocked off the surrounding holes, I could see the whole picture (although I used a lens to focus it.)

a similar phenomena can be viewed of an eclipse seen through the leaves of a tree. you see multiple images of the eclipse on the ground.

Here

actually it is a Camera lucida rather than C. obscura

:-)

Thanks for your comments everyone. Mollwollfumble, if you treat the light source as a spherical source rather than a point source, ray-tracing does suggest a full image at each point on the screen.

Thinking about this a bit further, I’m more and more convinced that there is a full image at each point on the screen, as Wocky’s experiments suggest. If you look through the hole from behind the screen, you’ll be able to see the whole slide. I still think I’d like to do the experiment myself.