Digital Pinhole Camera

There is a wealth of information about pinhole camera creation on the web, and I'm not going to try to recreate it here. I'd suggest checking out www.pinholevisions.org. This is one of the best single sites I've found. Check out the articles in the resource section. I relied heavily on this article: www.pinholevisions.org/resources/articles/Young. I also used a program called Pinhole Designer by David Balihar. It's for the PC, but a quick search finds a similar program for the Mac. I haven't found a Linux calculator, but there's a no-frills calculator on-line.

Step one was to determine what size pinhole to use. I started playing around with the calculator, trying different pinhole sizes.

I have two hard constraints, the size of the "film," which in my case is the CCD sensor in the webcam, which is about a centimeter square, and the exposure time, which is about 1/15 of a second.

My intuition was that a larger pinhole would be better, because it would gather more light. I guessed (correctly, as it turns out), that the biggest challenge in getting a digital

pinhole camera to work would be getting a bright enough image for the fixed exposure time of the webcam.

I was wrong about the larger pinhole. The nature of the pinhole camera is that different sizes of pinholes have different optimal focal lengths (the distance from the pinhole to the CCD). The bigger pinholes have a larger focal length, and it turns out that the old inverse-square law kicks in, and the light, although there is more of it, drops off even more by the time it travels the longer focal length, and ends up being dimmer. So, I needed a fairly small pinhole, and I needed a good amount of accuracy, so that I could get the best image quality. So I decided that rather than make my own pinhole, I'd plonk down $15 bucks for a precision laser-drilled hole. The smallest I could find was the .35mm hole from pinhole visions, pictured above. It arrives in pristine condition. For comparison, the picture above shows what it might look like if some idiot (me for instance) accidentally rolled over it with his office chair.

Here are the fixed elements of our system:

• CCD Sensor: 1 square centimeter
• Target wavelength: 650 nm [The wavelength of the laser pointer]

With a .35 mm pinhole, we get the following characteristics:

• Focal Length: 61.7 mm (2.4 inches) [Short Camera!]
• Field of View: 97° [A little wide,but the image drops of toward the edges, so may be fine)
• f-Stop: 176 [Can you feel the doom?]

The numbers aren't great, especially the f-Stop. I have my doubts that the laser will be bright enough to register on the CCD, but I'm willing to try.

For those interested in where those numbers come from, the formula is:

d = c * sqrt( f - L)

• d = diameter of pinhole
• c = Lord Rayleigh's constant = 1.9 [This number was derived empirically by LR]
• f = focal length
• L = light wavelength

Next up, building the camera.