This is the camera I had been talking about for six months. A wooden box, 20 inches by 16 inches on the back, with a five-kilo World War II aerial reconnaissance lens bolted to the front, built specifically to shoot portraits on black and white paper reversal. The whole thing came in at around £300 including wood, lens, fittings and paint. It weighs fifteen kilos. It took over most of my living room.
It is not finished as a project, and the results in this first round are not perfect, but it works, and I learned an enormous amount building and shooting it. If you are thinking about building something similar, here is everything I worked out, including the mistakes.
Why I built it
This started with an earlier video where I shot 11x14 black and white paper reversal in a cardboard pinhole box. The problem with that camera was the aperture. A pinhole is around f400, which meant exposure times of twelve or thirteen minutes for a reasonably bright landscape. Fine for landscapes. Useless for people, and people are what I like photographing.
So I wanted a camera that could take portraits using the same paper reversal process, which meant a proper lens with a usable aperture instead of a pinhole. The concept was simple: build a box out of wood, put a lens on the front at a fixed distance from the back wall, and that fixed distance sets the focus distance. Pose your subject at that distance and they are in focus. No focusing mechanism, no bellows, no rail. Just a box, a lens, and a known distance.
The lens
The lens drives everything in a build like this, so it is worth getting right first.
Tim at Negative Thinking loaned me a 360mm lens to start with. When I did the calculations, 360mm on a 20x16 back works out as a super wide angle, far too wide for flattering portraits. I needed something longer.
I watched eBay for a while and found a 20 inch (500mm) f5.6 Air Ministry aerial reconnaissance lens from the Second World War for about £200. No maker’s mark beyond the Air Ministry stamp, so I do not know much more about it than that. But two things made it right: it was long enough to give me a sensible portrait field of view on the big 20x16 back, and wide open at f5.6 it would give me exposure times short enough to actually photograph a person without them having to hold still for minutes.
I then designed the entire rest of the camera around that lens.
![PLACEHOLDER: the Air Ministry 500mm f5.6 lens, showing the Air Ministry stamp]
The interchangeable lens cone idea
The distance between the lens and the film plane (the back wall where the paper sits) sets the focus distance. I wanted some flexibility in that distance without building a bellows, which is complicated, and which I do not have the carpentry skills or equipment to do well.
So instead of a bellows, I went with interchangeable lens cones. These are removable cones of different lengths that mount the lens at different distances from the back, giving me a few fixed focus distances to choose from. Not quickly interchangeable, but swappable with a bit of effort.
The cones were 3D printed by Simon Forster at Forster UK, made to my specifications with his own structural additions to ensure they were strong enough to hold the four-kilo lens. I had two made: a longer cone giving roughly a 6 foot focus distance, and a shorter cone giving roughly 9 foot. I started with the 6 foot cone.
The cone system had a second benefit beyond flexibility. It was my insurance policy. If I got my measurements wrong on the box itself, I would not have to rebuild the whole thing. I could just adjust the cone dimensions to compensate. When you are building something this size with limited tools, having a way to correct mistakes without starting over is worth a lot.
I attached the lens to the cone, and the cone to the box, using bolts rather than screws, so the whole thing is more reversible and the wood does not get chewed up by repeated removal.
![PLACEHOLDER: one of the 3D-printed lens cones mounted to the front of the box with the lens attached]
Building the box
I am not a carpenter. I cut the wood on a Workmate and a garden table, sacrificing a spare piece of wood underneath the cuts, and then vacuumed the driveway afterwards because I had made a mess. This is a build anyone with basic tools can do. That is part of the point.
The construction is deliberately simple. The top and bottom panels are full size. All four sides screw up through the top and bottom panels into the end grain of the plywood. I used a drill guide to keep the screw holes perfectly straight so nothing poked through where it should not.
A few specific steps that mattered:
Cut the lens hole before assembly. I drilled and then cut the square hole for the lens in the front panel while it was still a flat, separate piece. Much easier than trying to do it once the box is built.
Paint the inside black after assembly, not before. I deliberately left the painting until the box was built so the black paint would seep into the corners and joints and help with light sealing. It is a bit trickier to paint a built box than flat panels, but the light-sealing benefit is worth it.
Hinged flap on the top rear for loading. The back needs a large hinged flap so you can get the paper in and out. I flocked the joint heavily and added flocking to all the edges so no light creeps in when the flap is closed, plus latches to secure it down and foam around the seal.
![PLACEHOLDER: the box part-built, showing the black-painted interior and the hinged rear flap]
Handles, and a centre of gravity lesson
This thing is too heavy to move without handles. My first instinct was to put them on the sides. My wife suggested front and back instead, which turned out to be the right call: with front-and-back handles I do not get a centre-of-gravity problem when I switch lens cones and change the weight distribution at the front. I added extra wood between the handles and the box to spread the load, because I was worried the short screws would rip straight out of the plywood under fifteen kilos.
Small detail, but if you build one of these, put the handles front and back.
Checking lens coverage
Before committing, I patched together enough paper to cover the whole back wall and checked that the 500mm lens covered the full 20x16 frame corner to corner. It did, with room to spare. The image actually spilled past the paper onto the painted base of the camera, which means I could have gone even bigger than 20x16 with this lens if I had wanted to.
![PLACEHOLDER: the projected image on the back wall, showing coverage past the corners of the 20x16 area]
The first shoot, with Tom
I tested it with a friend, Tom, before a long trip, so this first test was rushed and not ideal. I shot two sheets, one at roughly 12x16 and one at the full 20x16, working wide open at f5.6, controlling the exposure with my hand over the lens because the camera has no shutter. I metered with the Sekonic spot meter and added a fraction to the calculated time to allow for bellows extension (a one-second reading might get a second and a half).
The results taught me the two things that dominated the rest of the project. First, the depth of field at this format, wide open, at portrait distances, is paper thin. Genuinely millimetres. Getting both eyes sharp is very hard. Second, the exposures came out dark. I was underexposing, and I had not yet worked out the right extension compensation for the focus distance.
The 12x16 sheet was soft and a bit dark. The 20x16 sheet was sharper but still underexposed. Both were a start.
![PLACEHOLDER: the first 12x16 and 20x16 test prints with Tom, showing the softness and the dark exposure]
Developing prints this big
Processing 20x16 paper is its own challenge. I do not have the facilities or space to handle paper this size comfortably, so I adapted. I used a row of large trays for developer, stop, bleach, clarifier and fix, mixing two litres of each because one litre does not even cover the bottom of trays this big. For washing, I used a separate large tray of water that I periodically emptied into a plumbed-in sink.
The paper reversal sequence for these went: develop to completion, wash, bleach, wash, clarifier, wash, then room lights on, re-expose, develop, stop, fix, wash.
It is cumbersome and messy and you end up with enormous wet sheets of paper everywhere. But it works, and like everything else here it will get smoother with practice.
![PLACEHOLDER: the developing setup, showing the oversized trays needed for 20x16 paper]
The birthday shoot, and the struggle for a good exposure
After Tom, I got a second chance. It was my birthday, and my wife Nikki agreed to sit for me, which was the one thing I had asked for. The camera had taken over the living room. Three lights, blinds closed for some background separation, a 20x16 sheet loaded.
That one came back wildly underexposed again. There was an image there, and a nice one, but far too dark. I was not willing to end the project without one decent result, so I kept going across several more days and several more attempts, fighting the gloomy UK winter light the whole way.
By around the fifth test I was getting close. A 7-second exposure in what little winter sun there was started landing the exposure correctly. Focus was still not perfect, and with two young children as subjects there was inevitable movement, but the exposure was finally there.
![PLACEHOLDER: the final, correctly-exposed portrait, alongside one of the earlier underexposed attempts for comparison]
What I learned about the lens character
One genuinely fascinating thing emerged from the big prints. The image has an unusual sense of perspective. The subject feels wide, almost like a wide-angle rendering, but the background behind them is compressed and narrow, like a telephoto. That combination is unusual and quite striking, and it is a quality specific to this lens on this format. It is the kind of thing you only discover by building the odd thing and shooting it.
The other lesson: 20x16 prints are physically hard to take in. They are so big that standing in front of one, you cannot easily see the whole image at once. That is part of their appeal but it also changes how you have to think about composition.
Where the project stands
I ran out of time. I had to park the camera and head to Canada for several months, and there is no way a fifteen-kilo box camera was coming with me through customs. So the project is on hold until I am back, which by happy accident means picking it up again with far more daylight to work with. The grey overcast skies of a British winter were the single biggest obstacle to getting clean exposures, and the longer days should help enormously.
There is a lot more testing to do. I need more lens cones for longer focus distances (even at 6 feet, the 500mm is still wider than I would like for a tight head-and-shoulders portrait). I need to nail down the extension compensation properly. And I need subjects who can hold still better than a six-year-old and an eight-year-old.
But the camera works. The concept is sound. And for around £300 and a few weekends of fairly amateur woodwork, I have an ultra large format portrait camera that produces images nothing else I own can make.
If you want to build one, the headline lessons are these: pick your lens first and design around it, use interchangeable cones instead of a bellows if you cannot build a bellows, put the handles front and back, paint the inside black after assembly, and accept that the depth of field will terrify you. Everything else is just plywood and patience.