Last Update: 24 January 2016
Pinholes and UV Photography
Pinhole optics are the oldest known kind of imaging device, having been described as long ago as 500 BC. In the modern era, some artists have created works with them, but most photographers tend to dismiss them as children's toys or, at best, something suitable for a science demonstration. This is not without reason: even a cheap modern lens can produce an image that is many orders of magnitude more detailed than a pinhole image, allow control over the details of scene focus, and do it all in a fraction of the time. Moreover, many of the materials which were traditionally used in casual pinhole photography, such as black-and-white printing paper, are no longer easily available in the modern world. Nevertheless, there might be additional reasons to consider owning one of these devices.
First, even if one does not use a pinhole to produce finished photographs, it can be a useful and inexpensive bit of test equipment. Even the more expensive prefabricated versions for 35mm cameras and their digital descendents cost less than the typical used lens, as well as being much smaller and lighter (A real cheapskate can take a spare body cap, drill a pencil-sized hole in it, tape some thin aluminum foil on the inside, and poke a hole with the tip of a fine needle.) Its highly collimated light is very useful for mapping dust and debris on the sensor surface. In the ultraviolet domain, there is another application, which I have demonstrated elsewhere: because the pinhole transmits all wavelengths equally, it is an effective standard against which to measure the UV transmissivity of all other optics, often an important consideration when evaluating older lenses of unknown properties, when one does not have access to a spectrophotometer nor to a calibrated bandpass filter array. In conjunction with such an array, a pinhole can serve as a means to test the response of filter and sensor (independently from lens transmission.)
The question inevitably arises, given their low cost and outstanding bandpass properties, as to whether pinhole optics are a worthwhile option for producing ultraviolet photographs. The answer depends on what the photographer seeks, as well as the size of the sensing medium. Although a complete explication of the physics and optics of pinholes is beyond the scope of this post, reviewing some elementary facts can be helpful.
Pinhole Basics, in a Nutshell
Whereas a lens focuses incoming light, a pinhole does not; it merely selects a highly collimated subset of that light and uses that to project an inverted image on a flat surface. Therefore, a pinhole, at least in the geometric-optic approximation, does not have a focal length as such, only a projection distance which is just the distance from the pinhole to the sensing medium. A hypothetical point source at infinity will project a parallel-sided cylindrical beam through the aperture, ending in a circular spot called the circle of confusion. This spot is the same size as the aperture and this size effectively constrains the image resolution of the image. The spot size is not dependent on the projection distance, although the image magnification is. It is often said that everything in a pinhole image is equally sharp, but this is incorrect: as a point source approaches the pinhole, the beam behind the aperture becomes a diverging cone rather than a cylinder, and the circle of confusion grows accordingly, resulting in increasing blurring. In practice, if the subject distance is more than 20 times the projection distance, the circle of confusion will be within 5% of its ideal size and the image will be maximally sharp; closer than that, and there will be noticeable additional blurring. For a pinhole mounted 50mm from the sensing plane this works out to a subject distance of one meter. Geometrically, making the aperture smaller and closer to the sensor plane would seem to be a strategy to counter this, but diffraction effects forbid using an aperture smaller than a certain size; in reality, one cannot do close-up photography with a pinhole.
A pinhole is not, of course, merely a construct of geometric optics, but is subject to the wave nature of light as well. At certain geometries, it turns out that the beam behind the aperture converges slightly rather than being parallel, resulting in a slight maximum in resolution at certain combinations of projection and subject distance. The aperture acts, in effect, as a zone plate with one zone. Those who design pinhole systems usually try to take advantage of this "sweet spot" to maximize resolution, often using 550-nm light as a reference. Changing the wavelength or projection distance from what was designed may slightly deoptimize the system's performance (although the degradation may be small.)
Young has proposed the concept of "equivalent pixels" to define the performance of a pinhole setup. In this scheme the maximum resolution of a pinhole image is compared to that of a digital sensor of an equivalent number of pixels. For 35mm film, again using 550-nm light, the figure is 120 x 180 equivalent pixels. With 350-nm light and the same sensing area, the figure improves to 189 x 283 equivalent pixels. A smaller APS-C sensor manages only 118 x 177 equivalent pixels under the same circumstances, whereas an 8 x 10 film camera manages a nice, round 1,200 x 1,600. It is no surprise that almost all of the sharpest pinhole images have been obtained on large-format sheet film (including my own--I once took a UV photo with an 8 x 10 pinhole camera. Unfortunately, it was of no aesthetic merit and was discarded long ago. But it was sharp. Perhaps some day...)
The rest of this post will concern itself purely with digital UV pinhole photography, because that is the aspect of the genre most likely to be of any interest to modern readers. The digital photographer is at a clear disadvantage compared to the film photographers of yore, or even to the proverbial oatmeal canister with a sheet of black-and-white paper taped inside. Converted scanning backs may not exist. Oversized sensors (Mamiya, Hasselblad) would seem to be best-positioned to practice the craft; so-called "full-frame" sensors can be adequate; APS-C sensors are somewhat marginal, and one might as well forget using M43 gear for anything but test shots.
Choice of Subject Matter
Both the low resolution and the long exposure times involved in pinhole photography constrain what can be a subject and how it can be photographed. Botanical specimens are largely a poor choice due to their small size. Styles emphasizing subtle textures or a wealth of fine detail will not fare well. Bold, simple shapes are best, along with shapes such as trees where the human brain tends naturally to fill in missing detail. A pictorialist, soft approach works better. Some photographers have produced a charming, "toy" effect with pinhole photography, such as this one (not mine and not UV:)
The long exposure times mean that stationary subject matter is mostly required, and portraiture is very difficult (A related technique known as "pinspeck" photography offers much shorter exposure times, but is more esoteric and demanding and has far poorer dynamic resolution.) Scenes where long exposures would normally be used, on the other hand, can work well. The pinhole itself generates neither flare nor hot spots, but filters can and do generate such artifacts (the Baader filter exhibits an annoying yellow-green ring, which can be seen in some frames below.)
What to Expect on Attempting Pinhole UV
Exposure time: this will almost always be 30 seconds or longer, perhaps several minutes. As most cameras do not have an on-board exposure setting longer than 30 seconds, a wired remote unit with longer times is quite handy.
Viewfinders: SLR viewfinders are worthless with a pinhole, even without a filter (and keep the viewfinder shutter closed, or you will get nasty back-leakage fogging your image.) Electronic "live view" may or may not work. Often, if one does not want to be aiming blind, some sort of external auxiliary viewfinder or sighting device is useful. One does not want to crop these images more than absolutely necessary.
Workup: Cautious use of unsharp-masking can provide the illusion (if not the reality) of slightly greater sharpness--but don't overdo it. Expect to spend some time retouching dust specks. Even if you think your sensor is clean, using a pinhole will disabuse you of said notion.
Characteristics: Pinhole images are completely rectilinear and have no distortion. There is also very little chromatic aberration. Tonal range is generally good if executed well. The low modulation transfer function at certain spatial frequencies gives the images a kind of "smudginess" in addition to their blurred nature. Pinhole enthusiasts consider this to be a part of the charm of these images.
A Brief Annotated Gallery
All photos were taken with a converted Sony A900 camera, a Lenox Laser cap-mounted pinhole, and a Baader U2 filter. Display intent is BGR. Exposure data as indicated.
1. Test photos, with the pinhole mounted directly on the camera versus on a 69mm extension tube. 3 minutes @ ISO 200. Unretouched.
This is a demonstration that mounting the pinhole farther away than normal can indeed increase angular resolution (at the cost of smaller FOV.) Notice that the shape of the white statue is definitely more discernible in the bottom frame.
2. Untitled. 3 minutes @ ISO 500. Substantially cropped.
This photo shows the famous "Wolfe Angel" in Oakdale Cemetery in Hendersonville, NC, dusted with snow. The maroon star at the bottom of the right wing is not casually visible to the unaided eye. I am told that at one time, this marble statue was neglected and vandalized; later, it was restored and cleaned up. Perhaps the cleanup was not quite complete. It was snowing, and blurred water droplet on the filter is visible at lower left. My conventional photo of this statue may be viewed here:
3. "Upright Still." 30 seconds @ ISO 1600. Perspective corrected and somewhat cropped.
Cemeteries in the season when the leaves are off the trees seem to lend themselves well to this medium. The orange glow on the stela at right is the setting sun. The sky was blue and cloudless when this was taken. Cross traffic on a road in the background thankfully did not register.
4. "At Sunset." 30 seconds @ ISO 800. Cropped a bit on the right.
I find the pale color of the setting sun appealing, as well as the dark near-monochrome shapes in the foreground.
5. "Payload, Receding." 3 minutes @ ISO 200. Almost uncropped. Slight unsharp-masking.
This may be my best UV pinhole effort so far. On the original image, the lettering at the upper right of the rail car is legible. The orange color on the cars and the house at upper left is likely titanium dioxide.
"The Pinhole Camera: Imaging Without Lenses or Mirrors." Matthew D. Young. The Physics Teacher 1989, 648-655. Reposted at http://inside.mines....ng/PHCamera.pdf
"Pinhole Camera." Wikipedia. https://en.wikipedia.../Pinhole_camera
"The Pinhole Camera Revisited, or the Revenge of the Simple-Minded Engineer." Kjell Carlsson. https://www.kth.se/s...faa/Pinhole.pdf
Klaus Schmitt's remarks on pinhole UV photography: http://photographyof...earch?q=pinhole
A Quick & Dirty Transmission Test Using a Pinhole by Oldoinyo
Edited by OlDoinyo, 25 January 2016 - 05:03.