Silverbased

Today’s post is for folks who are just getting interested in pinhole photography—whether you’ve heard about it as a fun DIY project, or as a creative technique for producing evocative, dreamlike images.

Pinhole camera designs can be incredibly varied. The traditional scout-troop model was built from a Quaker Oats carton, exposing a single sheet of photo paper. But “single shot” pinholes are a bit inconvenient, especially as fewer people have access to a darkroom these days.

So lately I’ve been emphasizing pinhole designs using roll film—like my plasti-pinhole project and it’s siamese twin variant. (I also designed a 120-film camera for shooting 6×12 panoramas, a project you can find in the Best of Make book.)

Pinhole image on Fuji Acros 120 film, taken with a converted 1950s Argus 75 camera

But I’d like to give a little background that applies to all pinhole cameras, of any form. Then we’ll look at a method for fabricating the all-important pinhole itself.

Some Theory

So why does a pinhole camera work? Imagine a light-tight box, with a piece of film on one side and a tiny hole in the other. Each point on the film can only “see” one patch of the outside world, the one lined up with the pinhole—whether it’s light, dark, blue, red, etc. So an image of the scene forms upside down on the film.

With that idea in mind, visualize what happens if you move the pinhole closer: The angle from the film corners to the pinhole gets more oblique, and the camera takes in a wider view of the outside world.

In fact, the distance between the pinhole and the film is exactly equivalent to the focal length of a lens with the same coverage. Hence, it’s most informative to measure pinhole “focal lengths” in millimeters, just as with lenses. (Because a pinhole does not actually focus light, using focal length in this sense is technically a misnomer—but that’s the way most pinhole enthusiasts refer to it.)

So How Large is a “Pin” Hole, Anyway?

Up to a certain point, the smaller the hole you use, the sharper the image you get. But if you go too small, you run into a problem with diffraction—the tendency of light waves to fan outwards when they graze the edge of an obstruction (it’s a problem for lenses, too).

Thus, there is one hole diameter for any given focal length which gives the optimum possible sharpness. Historically, a number of great scientific minds labored to derive the proper formula to compute this. But today, you can just put your faith in a handy online calculator, like this one from Mr. Pinhole.

For typical cameras the best diameter works out somewhere between 0.2 to 0.5 millimeters—equivalent to an f/stop of f/100 to f/300. This definitely implies some long exposure times might be needed. But everything from infinity to inches away will be recorded with equal sharpness.

If you want a camera covering a particular angle of view, you have a choice between using a small piece of film with a short focal length, or building your camera using a bigger image format and a longer focal length. For example, a 30mm focal length on a standard 35mm film frame gives the same coverage as a 225mm focal length exposing an 8×10″ sheet (both show about 72° diagonally). But which is better?

As it turns out, the optimum pinhole diameter scales up more slowly than the focal length: The bigger you build the camera, the smaller the resulting f/stop, and the sharper the image. But eventually, ever-larger image formats can become cumbersome, costly, and impractical. (And besides, if it’s more sharpness you want—you could always use a lens!)

My own conclusion is that 120-film pinhole cameras offer a good trade-off between image quality and film-handling convenience.

How Do I Make a Pinhole?

The goal in fabricating a pinhole is to get one that is nicely circular, without ragged edges, and whose diameter you have at least roughly measured.

You’ll probably do best by piercing your hole in the thinnest possible material. If your pinhole more resembles a microscopic “tunnel,” oblique light rays will be blocked and you’ll get noticeable vignetting. While that effect can be interesting to explore, you’ll likely also see problems from light reflecting off the inner walls of the hole, degrading contrast.

While it’s fine to experiment with tinfoil or the side of a beer can, I settled on a method for fabricating pinholes in aluminum roof flashing (about 0.01″ to 0.02″ thick). I just find it more secure to fasten those flat, stiff sheets into whatever camera I’m building.

First, start by cutting squares of metal 50mm on a side. Why 50mm? Because it’s the same size as a 35mm slide mount—this may come in handy when it’s time to measure the pinhole diameter. Cut several extras, since there’s often a bit of trial and error in achieving your target pinhole diameter.

Place the metal square onto a piece of softwood, and tap a small dimple into the center. A ball-peen hammer works fine; however it might be easier to center the dimple if you hold something like the rounded head of a carriage bolt against the metal, then strike that with the hammer.

The dimple only needs to be high enough that you can selectively rub that spot against a sheet of sandpaper to thin it.

Press and sand the bump against fine sandpaper—320 or 400 grit works well. Here I show the sandpaper wrapped around a block of wood, to make it easier to hang onto; flexing the edges of the metal backwards will help avoid sandpapering your fingertips!

You want to rub with enough pressure that you’re definitely removing metal, but not so aggressively that you sand right through. After a noticeable flat spot has formed, it’s time to start checking your progress.

Gently press the tip of a sewing needle against the hollow of the bump. You are not trying to push it through yet—hold the needle by the sides to avoid applying too much pressure.

What you are testing is whether the metal is thin enough so that the tip of the needle telegraphs a tiny raised point through to the other side. If not, go back to the sandpaper and rub a bit more, then test again.

With your first test, you probably left a small pin-prick in the metal. Keep placing the needle in the same spot after that, to avoid inadvertently forming multiple pinholes.

Once the metal is thin enough for the needle tip to deform it, you should see something like this:

Take a couple of light strokes against the sandpaper to flatten the raised point. Now hold the metal up to a strong light—it’s possible that a tiny hole will show through already. But in any case, you’re getting very close.

Press the bump against a firm backing, like a telephone book. Replace the needle tip into the pin-prick you started, then give firmer pressure, pushing on the end of the needle as shown. You are not trying to push the needle through the metal—the diameter of its shaft is much too large. With gentle pressure you’re just trying to to push the point of the needle through.

Once you “see daylight,” go back to the sandpaper and lightly sand away any rough edges around the hole. Take the tip of the needle and very gently spin it in the hole to help round off any irregularities. Blow through the hole to remove any dust from the sandpaper.

Now It’s time to check if your hole looks clean, round, and what its diameter is.

It’s not necessary to achieve insane precision with your pinhole size. If you have nothing handy but a 10x magnifying loupe and a millimeter ruler, you can sort of “eyeball” whether your hole looks like one third of a millimeter, or whatever. Remember that a hole as far off as 70% or 140% of the desired diameter only means one f/stop of under- or overexposure, respectively. (This is within the exposure latitude for many kinds of film.)

But there are two easy methods to get a more precise measurement: Either use a slide projector, or a scanner connected to your computer. Today far more people have access to scanners than still own slide projectors—so I’ll describe that method first:

Simply put the pinhole metal into the scanner, and scan at the highest available resolution. (Only scan a small selection around the hole itself, to avoid ridiculously bloated document sizes.) Either scanning the hole on a flatbed document scanner or with the slide holder of a film scanner is fine—although a film scanner may offer higher resolution.

Some photo-editing software has a ruler tool allowing you to measure sizes directly. But even in a more basic program (like the old version of Elements shown here), you can still measure the hole. Drag out a selection which just barely encloses its image; then open the “info” palette to read off the size.

A highly-magnified scan of the pinhole; the selection is 107 pixels square. There also appears to be some gunk clinging to the hole that should be cleaned out

To get the most accurate measurement, I suggest going into the preferences and changing the default units to pixels. If you scanned the hole at 9600 dpi, and your selection is 107 pixels across, its diameter equals 107 divided by 9600, or 0.0111 inches. There are 25.4 millimeters in an inch, so that pinhole size translates to 0.28mm.

Personally, I find there is a lot of trial and error in achieving at a nice clean pinhole of the correct diameter. So I will often make a batch of pinholes at the same time, to get one good one. In that case, the scanner method becomes a little time-consuming.

So to me it’s worth setting up an old slide projector with a manual feeder, where I can quickly slap pinholes in and out, and immediately see how they look. This allows me to start with a hole slightly undersize, then nudge its diameter larger (by gently spinning the needle tip in the hole), quickly rechecking until I hit the target size. (Sandwiching the pinhole into a spare cardboard slide mount makes it fit in the projector gate more snugly.)

The key thing is to know that the opening of a standard slide mount is about 23×34mm. From the height of a slide image projected onto the wall, you can calculate the magnification; and from that, you can figure out how large the spot of light would appear from a properly-sized pinhole.

But I have an even simpler trick: Move the projector back and forth until a complete slide is projected at a size of 46 by 68 inches. At that magnification, every inch of the projected image represents 0.5mm at the slide mount. Thus you can read off pinhole sizes very rapidly (a ruler divided into tenths of an inch is helpful).

Measuring the projected image of a pinhole

If you discover you went wildly over your target pinhole diameter, start over with a fresh piece of metal, but try to press more gently with the needle this time. And don’t throw away the “bad” pinhole. Just write its diameter onto the metal using a permanent marker, and hang onto it somewhere. Building pinhole cameras is rather addictive; a day may come when you create another one needing exactly that diameter hole.

Figuring pinhole exposure times means knowing the equivalent f/number. This is simply the focal length of your pinhole camera, divided by the pinhole diameter. (But both need to be expressed in the same units, whether millimeters or inches.) It’s preferable to avoid f/numbers under f/100, just because the correct exposure in full sun will be a fraction of a second—not something most homemade pinhole shutters can time accurately.

When you need to calculate exposure times, few light meters will indicate f/stops all the way into the hundreds! So it’s helpful to know that f/128, f/181, and f/256 are exactly 6, 7, and 8 stops smaller than f/16. Find the indicated exposure time at f/16 and count off the steps to the correct (longer) time needed for your pinhole exposure.

Have fun!

In a pair earlier of articles, I showed how to gut a cheapie focus-free 35mm trashcam, and replace its lens with a pinhole.

One rationale for that project was that nowadays, it’s getting harder to buy and develop any film size besides 35mm. Pinhole cameras made for other odd formats are certainly fun, but sometimes require access to a darkroom to make them practical.

But one of the joys of pinhole photography is being able to try out bizarre, idiosyncratic camera designs: Weird-shaped frames, ultra-wide-angle coverage, or warped perspectives from curving the film plane.

The moderately wide coverage of the normal plasticam pinhole is interesting; and its standard 24×36mm frame makes developing the film at any lab easy. Yet compared to more exotic possibilities, it does begin to seem a little tame…

A Japanese woman on Flickr wanted to make a panorama-format pinhole camera. Her inspired idea was to take a plastic Holga, and saw it into pieces. She reused its film-supply and take-up-spool compartments, but replaced the middle section with a homemade “stretch limo” version: A light-tight box with film gate, pinhole and shutter. Genius!

A Holga uses 120 film of course; but she inspired me to consider doing something similar with 35mm film. (Aside from getting interesting widescreen framing, with pinhole cameras a larger format helps give a more detailed image.)

The way I build a 35mm plasti-pinhole, the original shutter is discarded; so it’s not the button on top that takes the picture any more. But the film-winding mechanism remains intact. As you wind, a toothed wheel allows 8 sprocket holes to go past, then locks; this yields the standard-width frame spacing. You still need to click the button on top to release the winding thumbwheel, before you can advance to the next frame.

But you do this independently from making the exposure. And I had an “aha” moment when I realized that if you clicked and wound twice between pictures, that standard mechanism would permit shooting double-width images: 24 x 72 mm!

But how to build the rest of the camera?

Here is my whimsically warped solution: I took two identical plastic trashcams, and sawed through them, exactly at the edges of their film gates. Then I glued and taped them back together “siamese-twin” style, to make a panoramic pinhole camera.

Two focus-free plastic 35mm cameras; $1.40 for the pair at my local thrift store.

A majority of trashy plastic models use curved film gates, to mask the deficiencies of their crummy lenses. But for this purpose, finding two matching cameras with flat film gates makes construction much simpler.

Some plasticams have their own “panorama” mask which you can swing into place. But all those extra parts would add more complications, so I shunned that style too. These two ultra-simple Bell & Howell trashcams turned out to be perfect.

I arbitrarily chose one camera for the supply-compartment half, and the other for the film takeup side. Then I disassembled both cameras and discarded all the unnecessary bits and pieces inside—lenses, shutters, springs, etc. It’s important to remove all stray metal parts before sawing into the camera body!

Discarding useless innards; black lines mark the approximate cut locations

Unlike in my standard plasticam-pinhole, I could not reuse the sliding lens guards as a shutter: The pinhole opening would not be aligned with either camera’s original lens position. So all those moving parts got tossed too.

Next I screwed the shells back onto the camera bodies, wrapped tape around both to hold their backs shut, and sawed through each one.

The crapcam types shown here include dummy weights glued into their bases (to lend an illusion of quality!) It was particularly tricky to avoid grinding the saw blade into those metal chunks—so be careful. An old hand-saw miter box is a great tool for getting a straight, square cut.

I sawed just inside the right edge of one camera’s film gate, and the left edge of the other. This still left some interior partitions standing in the way of the desired pinhole location, all of which needed to be cut away with a sharp knife.

Now, I wish I could say I used some sophisticated assembly technique to combine the two bodies. But really I just spooged the halves together with copious amounts of black silicone sealant—supplemented with much electrical tape. I was trying to fill all gaps where light might leak in, and keep the now-combined film gates as well-aligned as possible.

I also glued an aluminum bar across the two film-door halves, so the back would swing open and latch shut correctly as a single unit again.

A piece of aluminum sheet with an 0.2 mm pinhole went across the front of the camera body. Positioned only 26mm from the film plane, I knew this camera was going to give some wide-angle coverage! (Horizontally, it’s about 110°.) The pinhole size works out to f/128, for those of you keeping track.

For this camera I tried a new shutter idea, a design which has quickly became my absolute favorite. I will definitely be using it again for any future pinhole cameras.

I took a spare cable release I had lying around, and cut away the rotating barrel intended to thread into a shutter button. This uncovers enough extra length of the moving shaft to allow me to hot-glue it to a piece of thick black cardboard. Then the cable sheath is glued to the front of the camera body.

The moving shutter piece is outlined in red here. It has a cut-out which uncovers the pinhole as the cable release is pressed. A couple of scrap pieces glued around the edges guide the moving panel.

What’s wonderful about this design is that the cable-release’s own internal spring snaps the shutter closed again—or, it can be locked open indefinitely with the set-screw of the release. And there’s no jiggling the camera when you open the shutter.

The front shell of the camera needed a matching rectangular opening cut into it. Then finally, I screwed and taped the camera shell back together again. (There’s no particular significance to the metallic tape—it just hides several no-longer-needed openings in the camera’s front panel.)

The block of particle-board contains my usual homebrew tripod socket: A 1/4-20 nut epoxied into a hole in the bottom.

For a while I was jokingly calling this camera the “HaxPan,” in reference to Hasselblad’s multi-thousand-dollar panoramic XPan camera system. Of course, my camera covers a wider angle than even its 30mm lens (as well as saving a few pennies… )

Obviously no ordinary lab will know how to make prints from these crazy non-standard frames. But the negatives can be developed just like any other 35mm, and then scanned on any of the current inexpensive flatbed film scanners for further processing.

Making this camera was truly an experiment. There are still a couple of small light leaks at the joint between the cameras. And next time I would probably do a few things a little differently…

Placing the pinhole so close to the film does give extremely wide views—but you can’t see much detail at the edges, because the light fall-off is so extreme. And a 110° angle of coverage is so hard to visualize without a proper viewfinder that I found framing to be quite hit-or-miss.

So next time, rather than placing the pinhole so far back, I would build up the light-tight inner compartment a little deeper and use a slightly longer focal length.

But that’s the great thing about pinhole cameras: The cost of experimenting is low… and the fun of blazing new ground is priceless.

Happy hacking!

Sample photo from the double-width 35mm pinhole. Note small light leaks at the seam between camera bodies. More samples here.