Programming Project #4

Building a Pinhole Camera
15-463: Computational Photography
Author: Stephen Tawa (stawa) and Bill Ge (bge)

Overview

For this assignment we made a pinhole camera and experimented with different aperture sizes.

Our Design

Our box was slightly larger than a shoebox. It was mostly lightproof but there were a few spots that required duct tape. The camera and pinhole were close to the center of their face. To use different aperture sizes, we cut a large (> 5mm) hole in the box and covered it with black construction paper containing the pinhole. The formula 1.9 * sqrt(f * lambda) suggests a pinhole size of .6mm, so we tried .6mm, 2mm, and 5mm.

Some pictures of our setup:

Results (flipped vertically)

Scene 1, .6mm pinhole:

Scene 1, 2mm pinhole:

Scene 1, 5mm pinhole:

Scene 1, .6mm pinhole:

Scene 1, 2mm pinhole:

Scene 1, 5mm pinhole:

Discussion

The main difference between smaller and larger pinholes was how to capture an appropriate amount of light. With the smallest pinhole, we needed to use very long exposure times (usually 60 seconds). With the medium and large holes, we had to shorten the exposure time to prevent the images from washing out. 15 to 30 seconds usually worked. If the largest pinhole was pointed directly towards the sun, it was necessary to create shadows. Otherwise, the image would become washed out even with short exposure times. As expected, the smaller pinholes generally produced sharper images.

Some more images:

Discussion

It was harder to create proper lighting conditions inside. For the last image, we used a flashlight to generate additional light. These pictures all used the .6mm pinhole with a 60 second exposure time, since this resulted in the sharpest images.

Bells and Whistles

Light Painting

Fail attempt at stick figures...

Alec chillin with a stick figure

KAMEHAMEHA

This is how I travel home

Using a Lens

Here is the new setup with a lens:

Some results with the lens:

Discussion

Using the lens creates circular images that do not fill the whole sheet. We had to cut another hole farther away from the camera, so we experimented with different angles. We leaked light in the second image when we changed the camera angle, but the third image was successful.