Lenses are (of course) a critical component of any photographic system. The multispectral photographer has two main choices in this area: either purchase one or two very expensive multispectral lenses, or use regular lenses and accept the tradeoffs.
Professional Multispectral Lenses
For those with deep pockets who want to pursue the former option, the current "belle of the ball" is the Jenoptik (formerly Coastal Optics) CoastalOpt UV-VIS-IR 60mm f/4. This quartz/fluorite Nikon F-mount lens is said to be truly superb for human-visible, near infrared, and ultraviolet photography. It retails for about US$4500, so I will probably never even be allowed to hold (let alone think about purchasing) one.
A slightly less-expensive alternative is Nikon's own UV-Nikkor 105mm f/4.5 (no longer produced under the Nikkor brand, but available from Tochigi Nikon according to Bjørn Rørslett). This short telephoto model is a mainstay of professional photographers who shoot multispectrally, and (like the Jenoptik lens) spoken of glowingly. It had a list price of US$3000, and typically sells second-hand for US$1500-US$2000.
That's about it. Like Henry Ford's range of colours, you can have any focal length you want, as long as it's 60mm or 105mm. NASA commissioned at least two 55mm UV-Nikkor designs directly from Nikon, but they were produced in extremely limited numbers, and so command very high prices on the rare occassions when they appear on the secondhand market. For example, in December 2009, a collector listed a NASA-customized Nikon F3 with one of the later 55mm UV-Nikkor lenses on eBay with a selling price of US$69,000.
Standard Camera Lenses
Beyond the ability to pass ultraviolet light unimpeded (conventional glass filters it fairly effectively), the main selling point of both of these lenses is that their designs are compensated for the multispectral equivalent of chromatic aberration. This is an optical effect which is related to the behaviour of a prism; a lens will alter (e.g. focus) the path of light at varying angles depending on the wavelength of the light:
The basic theoretical model that applies to simple optics. Real-world imaging systems (including camera lenses and telescopes) incorporate more advanced designs that minimize this effect within the spectral range for which the equipment was engineered.
Conventional camera lenses use designs that generally do a good job of nullifying this effect as far as it applies to human-visible light, but this does not apply to infrared or ultraviolet-A frequencies. This is why on older lenses you'll often see a red dot on the focusing ring which indicates the offset at which near infrared is in focus.
The two lenses above use materials (quartz and fluorite) which not only transmit the entire ultraviolet-to-near-infrared range, but which focus most or all of that same range within nearly the same distance from the lens. This means that the focus can be set using the viewfinder, then a UVA or NIR filter placed in front of the lens and a picture taken at that time will be in focus even though the focus was set using a different part of the spectrum. Obviously this is an incredible time-saver compared to the method below. Note that only the CoastalOpt lens design has this ability for the entire ultraviolet-A-to-near-infrared range. The UV-Nikkor must be refocused for near-infrared imaging, but not ultraviolet-A.
As a shallow-pocketed hobbyist, I've learned to make do with standard lenses. This is certainly a viable option, but a couple of caveats apply. First, as noted above these lenses are not corrected for the different effective focal lengths of frequencies beyond human-visible light. Using the diagram above, you can see that the expected theoretical result would be something like this:
|Effective Focal Length With Uncompensated Lenses|
The general idea of how standard lenses will behave when used for multispectral photography. They are "farsighted" in the near infrared, and "nearsighted" in terms of ultraviolet-A light.
...and in fact, this is exactly the result in many applications. For landscape photography, a more distant area will be in focus in the near infrared exposure than what you would see through the viewfinder, and an area closer to the camera will be in focus in the ultraviolet-A exposure. At wider apertures, this often means that "infinity focus" is not actually focused on infinity in the ultraviolet-A shot. Because most conventional lenses (sensibly enough) do not allow focusing beyond infinity, landscape ultraviolet-A shots require a very narrow aperture; I generally use f/22. This effect is also important to keep in mind for near infrared photography, because it's easy to assume that a relatively distant feature will be in focus, when in fact it's still too close for the "nearsighted" infrared response to capture sharply.
However, there is also a somewhat unintuitive (unless you are well-read in optics) effect which can take place instead, particularly with close-range (non-infinity-focus) photography: both near infrared and ultraviolet-A may require focusing closer than the setting for human-visible light. This is due to a quirk of "achromatic" lens designs. Nikon's MicroscopyU has a couple of good illustrations (about halfway down the page and about 2/3 of the way down the page). You can see that with an achromatic design, the effective focal length actually increases towards both ends of the spectrum.
Whichever effect takes precedence, a lot of patience and trial-and-error is required. See Taking Multispectral Pictures for details.
The second issue that most people notice when using regular lenses is that much longer exposures are required for ultraviolet-A light. I generally "lose" between 6 and 9 stops of light sensitivity, depending on the lens and lighting conditions, and that's assuming I'm using the same aperture for human-visible and ultraviolet-A exposures (which is only infrequently the case now that I'm more experienced). Unless you are shooting in broad daylight and have a camera with LiveView, you're probably not going to be taking any effective ultraviolet pictures of subjects which move faster than geological features and plants.
The final potential headache of using standard lenses regards "hotspots". These are most frequently exhibited when shooting near infrared, but can crop up with ultraviolet photography as well, depending on the lens. I don't think anyone has truly answered the question of what specific combination of lens design and lighting conditions cause them, but they are a cause of great aggravation and annoyance. Using an umbrella or other sun-shade can help (see Other Tools), and their intensity is inversely proportional to aperture width, but some lenses are virtually unusable because of this problem.
As in most cases, Bjørn Rørslett's site is an invaluable resource, as he has reviewed numerous lenses and frequently comments on their performance in the near infrared and/or ultraviolet.
Like Mr. Rørslett, I've usually obtained the best results when using older lenses. I believe the working theory is that newer lenses have more optical coatings designed to reduce artifacts within the human-visible spectrum, but that these same coatings impede the performance outside those boundaries. I began using Nikon's old Series E lenses almost exclusively, but have since adopted several even older models, and have been greatly pleased with the results.
Many of Nikon's older lenses are available very cheaply, and they tend to be much more solidly-constructed than models from the last decade or two. Nikkor lenses from the 60s and 70s were built entirely from metal and glass, and feel like they will last forever. If you are interested in going this route, be aware of the difference between "AI" ("automatic indexing") and "pre-AI" or "non-AI" lenses. In 1977, Nikon revised their lens mount design, and lenses of the earlier type cannot be mounted on a modern DSLR without being modified. At the time, Nikon offered factory conversion, and so there are some lenses on the secondhand market that were thus modified. Most of them were not, and so much be altered by a third party if they are to be used on a DSLR. I've had two lenses converted by Brian Williams, and he did an excellent job very quickly. He was also kind enough to remove some internal dust from one of the lenses which the seller I bought it from hadn't disclosed. The other big name is John White, although I haven't had any experience with him myself.
So far, my favourite results have been obtained with a couple of Nikkor-Q lenses: a 200mm f/4 (factory AI-converted), and a 135mm f/2.8 (converted by Brian Williams). These two lenses combined (including the AI conversion of the latter) cost me less than any new lens, even the all-plastic Nikkor 18-55mm f/3.5-5.6 (which incidentally is no slouch in terms of multispectral performance). You can see some examples shot using these two in The Colours of Autumn.
The aforementioned Series E lenses have also served me well. Every photo taken in my 2007 drive (see Drives) was shot with a Series E lens, and many before and after as well. I have not gone to the same lengths as Bjørn Rørslett, in that I use them in their stock form, as opposed to polishing off the lens coatings.
Once I've had a chance to compare the numerous options, I plan on publishing detailed reviews, but for now I once again recommend Bjørn Rørslett's lens reviews for anyone interested in this approach.
|1.||If you are hoping to find one in less-than-ideal condition for less, best of luck to you. I've seen them sell for $1500 on eBay even when the seller had no idea whether it was even usable anymore.|
|2.||Nikon's use of 105mm and 55mm focal length designs for the UV-Nikkor lenses can easily give the (false) impression that these were simply rebadged or modified versions of the Micro-Nikkor lenses which were produced in the same focal lengths. This is not the case; not only are the elements made from specialized glasses, but the actual optical designs are completely different. See and for a comparison of the 105mm lenses (6 elements in 5 groups for the UV-Nikkor versus 5 elements in 3 groups for the Micro-Nikkor). See and for a comparison of the 55mm lenses (3 elements in 3 groups for the UV-Nikkor versus 6 elements in 5 groups for the Micro-Nikkor).|
|3.||Some examples are achromatic and apochromatic designs. This page has examples from the astronomical world, and this one applies specifically to microscopy.|