UV photography  

In addition to visible light (i.e., wavelengths ranging from 400 to 700 nm) and infrared, sunlight contains UV-A (or near ultraviolet, NUV, 320-400 nm), UV-B (320-280 nm) and UV-C (280-200 nm) light. UV-B is absorbed by most materials normally used to manufacture lenses, and digital camera sensors also possess low sensitivities in these ranges. UV-C is rapidly absorbed by atmospheric oxygen, and is largely irrelevant to photography. Camera sensors are moderately sensitive, instead, to UV-A. Special lenses can be designed for UV-A and UV-B photography, and at least some of the general-purpose photographic lenses also transmit rather low amounts of these wavelengths. On this page I discuss several of these general-purpose lenses that turn out to be suitable for NUV photography. Therefore, it is possible to use digital cameras and general-purpose lenses to take pictures in the near UV range, at least for static subjects that allow relatively long exposures. Among Nikon DSLRs, the D70 and D70s are reported to be particularly suitable for UV photography. Canon and other brands are said to be unsuitable. I discuss here a Panasonic G3 that I modified for multispectral (including UV) photography with very good results.

UV photography normally requires a filter that blocks longer wavelengths than UV, so that they do not contribute to pictures. When we talk about UV filters, usually we mean filters that are opaque to UV and transmit longer wavelengths. These filters are more properly called UV-cut filters. This is the opposite of what we need for UV photography. What we want is UV-pass filters. A special problem with digital cameras is that their sensors are very sensitive to near infrared (NIR), and any NIR that leaks through a UV-pass filter will usually spoil the recorded image. I discuss several suitable and unsuitable UV-pass filters at length on this page (and other pages linked from there).

My early trials with near-UV photography were not much successful, because of - very large - technical problems that stack up to make the odds almost impossible. Unlike IR photography, which basically requires only a cheap, easy-to-find filter, getting the proper equipment for UV photography is likely to be both expensive and time consuming. Essentially, cameras, digital sensors, lenses and ordinary filters are all designed to cut out UV light, and you have to swim upstream in order to obtain odd bits and pieces that, by accident or design, work in the opposite way of what is normally desired.

A few (generally old and single-coated) general-purpose lenses do transmit usable amounts of UV. However, virtually all of them suffer from a problem not frequently discussed in photography forums. These lenses are corrected in the visible range, but exhibit varying amounts of focus shift between visible and UV. This phenomenon is comparable with the focus-shift between visible and IR, which is a much more widely known phenomenon. In practice, this means that focusing in visible light produces an out-of-focus picture in the UV or IR.  Many lenses have a red dot on the focusing scale, used for manually compensating for focus shift in IR photography. There is no guarantee that focus shift in the UV is in the same direction, or by the same amount, as in the IR.

I attempted several of the commonly discussed "fixes" for near-ultraviolet photography, (i.e., attempts at using and/or modifying lenses not designed for this purpose), with scarce success. Not being able to afford a UV Nikkor 105mm, which would have removed most problems, I tried, for instance:

  • Removing the coating on the front lens element in a Nikkor E 35mm f/2.5 and a Nikkor E 100mm f/2.8. These lenses are described as the subject of a similar treatment on another web site. I did this by rubbing the lens surface with a damp cotton cloth dusted with cerium oxide polishing powder. This substance is very hard, and used for polishing gemstones. Also the lens coating is very hard, but rubbing by hand for a couple of hours did remove most of the coating. An unexpected problem was that the lens material is much softer than the coating, so the lens surface turned slightly bumpy on a microscopical scale. The polished lenses did transmit fair amounts of UV (see example here), but also turned into interesting examples of soft focus lenses (the effect is slight, but interesting for portraits, and subtly different from using a soft filter). However, a soft focus lens is not what I wanted. Removing the coating by chemical treatments has been attempted by other photographers, albeit with no reports of success.
  • Using some enlarger lenses that are color-corrected into the near-UV. The EL-Nikkor 63mm f/3.5 is often recommended for this purpose. Perhaps because of this fame, it seems to be hoarded by collectors, and is much scarcer and more expensive than other EL-Nikkor models and focal lengths. However, my tests show that eight other EL-Nikkor models are roughly as good for near-UV photography. (see also my comparison of 63 mm EL-Nikkor models in near-UV and photomacrography). Probably, all EL-Nikkors are approximately as good for this purpose (see also independent information here, or the whole blog here).
  • Using normal lenses with a UV-pass filter and enormous exposure times. Eventually, some UV becomes detectable, but usually buried in noise.

In the end, my current, satisfactorily working solution for UV photography involves a few unusual pieces of equipment (not counting UV light sources, and various adapters, bits and pieces needed to make this equipment work together):

The UV Rodagon and Schuler filter were used, for example, in the above pictures, which show the UV nectar tracks (darker areas of the petals) that attract insects to flowers. These features are completely absent in the visible and infrared ranges.

As UV-only sources, initially I tried both "black light" fluorescent tubes (of the type that can be screwed into an ordinary socket for a light bulb) and UV LED banks (which can be mounted in 12 V sockets for halogen lights). The latter seem to produce more blue and indigo light and less UV than fluorescent tubes (at least in the models I tried). LED banks can be battery-powered or fed with a 12V DC power supply, but they draw so little current that the 12 V solid-state transformers used for halogen lights sometimes are fooled into thinking that no lamp is present, and refuse to provide a current. Ordinary incandescent light bulbs (especially those of higher power and higher Kelvin temperatures) do emit potentially usable amounts of UV, but they also emit a lot more IR. I tested additional UV sources here and here.

There are specialist continuous light sources capable of producing copious amounts of UV (e.g., for adhesive curing and fluorescence microscopy). The tube of electronic flashes does produce a lot of UV, which normally is filtered out by a yellowish coating on the outer surface of the tube (sometimes the filter is coloured glass used to make the tube, and therefore is not removable). The plastic exit window of flashes also cuts UV, and sometimes has a distinct yellowish colour that indicates the presence of UV-absorbing materials. There are a few specialist electronic flashes designed to provide unfiltered UV, but they are scarce, and probably mostly discontinued (this includes the Nikon SB-140). In principle, one might attempt to convert a general-purpose electronic flash for UV-use by removing the tube coating and plastic exit window. Studio strobes, which use naked tubes, might also be modified if the tube coating can be removed. There are also uncoated replacement tubes for studio strobes (in fact, they are cheaper than the ones coated with a UV-cut layer). Cheap, amateur-level studio strobes made in the Far East normally use uncoated tubes. These strobes are neither durable nor especially powerful, but they may be the best compromise between price and quality for studio UV photography. (See here and here for a more updated discussion of studio flash units for UV photography.)

A word of caution when modifying electronic flashes: their capacitors store a potentially deadly high-voltage charge even when the power is turned off and power cords are removed. The capacitor may remain charged for hours or days. There may be a dangerous charge left in the capacitor even after manually triggering the flash. This applies also to battery powered units. Make sure you know what you are doing.

UV damages the eyes (especially UV-B and UV-C, but UV-A is by no means safe), so you should not stare into UV light sources. Even if the sources are not powerful, exposure to UV is compounded by the fact that the eye's iris does not contract in the presence of UV, thus increasing the effective irradiation of the eye. It is also a good idea to switch off the UV sources when not in use and to wear protecting glasses that cut UV radiation, to prevent long-term exposure to indirect UV during photography sessions. Most plastic glasses absorb UV, and glass prescription lenses can be ordered with a colourless UV-absorbing layer.

UV photography is very different from photographing UV-excited fluorescence in the visible range. Fluorescence is the emission of longer wavelengths than those of incident light. Thus, illumination with UV radiation can cause the subject to fluoresce by emitting visible light. This is, for all practical purposes, photography in the visible range, and in fact it does not require special lenses and may be carried out with a UV-cut filter on the lens. The above pictures are an example of UV-excited visible fluorescence. Some of the markings on a banknote, which are almost invisible in ordinary light (left), fluoresce strongly when illuminated with near-UV light (right). Visible light illumination was used in both pictures (in addition to UV illumination in the second picture).

The above picture shows visible fluorescence of a white cardboard sheet illuminated by the UV fluorescent tube visible in the picture. The fluorescence in the picture is much brighter that the surface of the UV bulb. This picture was taken without additional light sources. This example shows that, even with a UV-only light source, often it is necessary to use a UV-pass, visible-cut filter on the camera, in order to eliminate visible fluorescence of the subject that may overwhelm its actual UV reflectancy. Some UV fluorescent tubes do emit also large amounts of IR.

Some people report "seeing" UV light. Most of these people do not possess unusual capabilities (there are documented genetic traits that do produce minor differences in spectral response of the eye photopigments). Instead, those who report seeing UV at improbably short wavelengths may be experiencing UV-induced fluorescence of parts of their eyes, which causes a stimulation of the retina in the visible range. The most common result is a perception of "haziness" in the presence of UV.

UV may be recorded in slightly different ways by the solid-state sensors of ordinary cameras. The longest UV wavelengths are generally recorded mostly as blue/indigo, and shorter ones, in order of decreasing wavelengths, mostly as red/gray, yellow, green and blue/cyan. In principle, this allows the diferentiation of multiple bands of near-UV. I do not know the reason for this difference, but it may be a combination of different transparency of Bayer filters on the sensor chip, fluorescence of these Bayer filters, and spectral sensitivity of the photo-sites. The solid-state sensors of ordinary cameras seem to be poorly sensitive to UV at wavelengths around 320 nm and shorter. CCD sensors seem to be more sensitive to UV than CMOS sensors, but exceptions are known. Older cameras like the Nikon D70 are often best for this application, but some of the newer cameras (e.g., Panasonic Micro 4/3 models) are also excellent. Relatively recent models like the Nikon D90 are said to be essentially useless for UV photography, even after modification, but the Nikon D700 seems to be quite good. These differences may be due to the varying behavior of the transparent windows covering the sensor chips and of their antireflection coatings, as well as to intrinsic chip properties that include UV transmission by Bayer filters and microlenses.

Special sensors for UV videocameras have no Bayer filters, often no microlens array, and are either covered by a quartz window or completely "naked" (which leaves the sensor chip very vulnerable to accidental damage). These special sensors can record UV wavelengths as low as 200-240 nm and are often used with UV excimer lasers commonly employed in the semiconductor industry and in biomedical science.

page hits during the last year
(whole site)

Copyright (c) Enrico Savazzi, 1990 - . This page was created in 2009 , last updated July 1, 2013
(last update refers to any change of contents, styling and/or web host)

This site is ad-free. If you see any ads here, they are added by your ISP, or by spyware on your computer, or you are visiting this site through frames of another site.