If you're one of the few, the blessed few, the regular readers of this humble blog, you've probably noticed I post infrared photos here pretty often. Ever since I realized you could take infrared pics with a digital camera, I've been sort of mad for it. Once you figure out that your camera can see below the visible spectrum, it's only natural to wonder whether it can see above the visible spectrum too. Or at least it's only natural for me to wonder.
While infrared photos tend toward dramatic landscapes full of treees and puffy clouds, ultraviolet photography is dominated by closeup/macro shots of flowers. Looking at flowers in UV can reveal details you can't see with the human eye, but which bees apparently can see. I remember seeing example photos of this as a kid and wishing I could see in ultraviolet too, or at least take pictures in UV and see the invisible that way. Plus it gives me an excuse to post even more photos of flowers, which are a perennial staple on this blog. Or at least it'll be a good excuse in a few months, once flowers start to bloom out again.
If you search the interwebs for a few minutes, the short answer is that digital ultraviolet photography is either a.) impossible, or b.) prohibitively expensive. The main problem is that ordinary lens glass isn't entirely transparent in ultraviolet frequencies, and becomes increasingly opaque as the wavelengths get shorter. Visible light cuts off around 400 nanometers. Glass, we're told, is mostly opaque by 350nm (the exact value seems to depend on who you talk to). So if you're going to be a serious Real Man photographer about it you'll need to shell out around $5000 for a quartz lens, which transmits down to around 200nm. If you want to go even lower than that, you'll need rather heroic measures, as oxygen becomes opaque around that point. You'll need an evacuated bell jar for your macro shots, large enough to hold your camera. And for landscape work you'll most likely require a spaceship or something. You'll also need a source of short-wavelength ultraviolet light. If you haven't yet left the earth, you'll need your own source of UV light below around ~315nm or so, because that pesky, naughty ozone layer is blocking the sun's UV. Below a certain wavelength, it's always night on earth, effectively, and everything is black (i.e. highly UV-absorbent). The ozone layer is rather handy for the whole "not dying" thing, but photographically it's a bit inconvenient, I guess. Effectively there's a sort of law of diminishing returns in effect here, under which you chase shorter and shorter wavelengths at ever-increasing expense, as the available natural light decreases rapidly. In short, I concluded that a.) I can't afford a hyper-specialized $5000 lens, and b.) I have no desire to mess around with dangerous UV-C lighting, no matter how fabulous the results might be, and c.) I wanted to see if I could pull this off anyway, regardless of what the self-styled experts think.
There is one thing I seem to have done differently than the examples I've seen on the net, in that I'm using a cheap point-n-shoot digital camera instead of an expensive DSLR. As with IR, it seems to be the case that the better camera you have, the better job it does filtering out UV. This is perfectly understandable; 99.9% of camera users don't want UV around messing up their photos, 100% of the time. I always suspected my lil' Canon A520 was detecting UV, as the sky tends to get overexposed and washed out at the slightest provocation. So I think if you want a camera to take digital UV photos, worse is probably better. I haven't tried the UV filter on my super-cheapies, the VuPoint and the Jamcam, just yet, but I expect I'll get around to it fairly soon, and you'll see the results here if they're remotely interesting.
Unlike IR, I don't know of any way you can play with UV without spending any money. Before burning $40 on a filter, I first thought I'd do a quick smoke test to see if it was worth trying. I purchased a black light bulb around Halloween, and I thought, OK, obviously this isn't a pure UV filter since I can see through it, but it has to transmit a disproportionate amount of UV. Hopefully it skews things enough to make UV effects detectable, at least. So I used the bulb at a filter, and pointed it at a partially open window. Normal glass doesn't do a great job transmitting UV, so I figured, ok, show a partially open window, so only half the frame is glass, and see if there's a noticeable difference. See which is brighter. Sure enough, the side without glass was substantially brighter, indicating that something the camera detects is being partially absorbed by the glass. This effect isn't noticeable in visible wavelengths, and I don't think it's noticeable in IR either. Obviously the glass is absorbing something and there aren't really that many candidates out there.
Assuming you can detect UV light, your next issue is being sure you're blocking out everything except UV light. The infrared equivalent was easy -- you just need a couple of strips of developed, unexposed film, tape 'em together and you've got a crude IR filter to get you started. As far as I know there's no equivalent in the UV field. Either you shell out around $40+ for a filter, or you don't. You can try the semi-UV trick I did above, but that's not likely to satisfy anyone, certainly not the hardcore types out there. I took a few pics with the bulb-filter, and they seemed to sort of capture the spirit of UV photos, if there is such a thing, but they weren't quite pure enough. So I broke down and bought the UV filter, a B+W 403. The B+W 403 has an unfortunate trait in that, although it blocks visible light quite well, it does transmit a bit of infrared light too. So I also bought a proper IR-only filter for comparison (Hoya R72), another $40. I figured I'd use it a lot anyway just for IR stuff, so it's really not that huge of an extravagance, I guess. I also needed a 52mm filter adapter for the digicam, Canon part number LA-DC52F, another $18. And if you count the price of a couple of black light bulbs, I've spent a bit over $100 on this project so far. Which is not an insubstantial amount, but to me it's affordable, and $5000 absolutely isn't.
Perhaps you've noticed the slightly defensive tone of this post. I've done a bit of research around the net, and one thing I've noticed is that any mention of ultraviolet photography tends to attract hordes of arrogant, pompous, insufferable blowhard windbags, and they all want to be the first to ridicule you for not buying a $5000 quartz lens like you're supposed to. No quartz lens, no admission to the exclusive UV photo treehouse. Them's the rules. Possibly they'd be better served by spending that $5k on anger management classes, but to each his own, I guess. Want some examples? Read the user comments to these pieces at Instructables, Make, and LifeHacker. Yikes! The worst part is that I get the distinct impression many of the angriest commenters out there have never actually tried it themselves. They simply read it was impossible, somewhere out on the interwebs, and decided the net needed yet another Meme Enforcer.
So I figured that if I wanted to demonstrate it was possible for less than the price of a used car, I ought to be somewhat painstaking about it. I don't want to claim it if I can't prove it beyond a reasonable doubt. I had a cheap digital point-n-shoot camera from 2004, a UV (and a little IR) filter, and IR (and no UV) filter, and a fluorescent blacklight bulb, which emits plenty of UV and almost no IR. And I did this at night, so the sun wouldn't mess with my results. I'm not trying to make art today, I'm not trying to argue it's an efficent process or anything. I'm simply trying to prove it's possible. If that requires artificial conditions, so be it.
The logic: I have a camera that detects visible light, some UV, and some IR, and that's apparently all. So to my knowledge, any light it sees must be either visible, UV, or IR. I have a known source of light, which puts out a great deal of UV, and very little IR, I take a photo with the UV filter, and it comes out fairly bright. But that could be due to the filter's IR leakage. To eliminate the possibility that it's IR, take the same photo of the same UV source using the IR-only filter. Note how much dimmer it is. The original photo was pitch black, and I had to enhance it a great deal to make the blacklight bulb visible. I think IR is out. Then look through both filters at the light source. Note that both transmit a very small amount of dark red visible light. The B+W 403 filter is visibly darker, and yet takes a brighter image of the bulb than the R72 filter does. Certainly the 403 doesn't transmit enough visible light to account for the rather sharp images it allows. UV is the only remaining candidate, by process of elimination.
So the top photo is the black light bulb with the B+W 403 filter on. Below it is the same black light bulb with the IR filter on, and the IR photo is heavily enhanced so you can see the bulb. The original was basically pure black, even though it was a longer exposure with a wider aperture. Fluorescent bulbs don't put out a lot of IR, so what you're seeing there might even be ambient light from elsewhere in the room reflecting off the bulb. Below, with the 403 filter again, a regular fluorescent bulb.
One disappointing thing is that I probably won't be able to use the ol' Holga for UV work, due to the plastic lens. Not all plastic is created alike when it comes to UV transmission, and I understand that acrylic (aka Plexiglas (tm), aka Lucite (tm) ) actually does a better job than glass in some respects, transmitting UV down to ~300nm. Other plastics tend to have a cutoff around 400nm, making them exactly as bad as the human eye in the UV department. And sadly, I hear the Holga's lens is polycarbonate plastic, not acrylic, and polycarbonate cuts off around ~400nm, making it no better than the human eye for this sort of thing. Oh, well. I should point out that I've never tried the acrylic lens option myself,
A fun thing with digital is you can try to prototype something you want to do on film without wasting any film. Film is, if anything, more sensitive to UV than digital is, so if digital works, film probably will too, and digital kind of looks like it works. I hear the very best thing to use for UV is old-sk00l, orthochromatic, blue-sensitive-only film, if you can find it.
Below, another pair of UV vs. IR shots of the same subject. The UV shot is just semi-UV, since I was still using the black light bulb as a filter at that point.
Resources:
- The UV section at Naturfotograf shouldn't be missed. Mr. Rorslett is one of the lucky few to own a quartz UV-Nikkor lens, so the site is a bit quartz-centric, but hey. Use it if you've got it. This site may be where the quartz-only meme got going, although I think this was probably unintentional.
- There's another good source of info at RMIT University in Melbourne, Australia. It focuses on using UV for medical photography, but the technical side should be handy for anyone with an interest in the subject.
- More info, with an enjoyably technical bent, at Beyond Visible.
- In particular, here are pages on filter transmission at RMIT and Beyond Visible.
- More digital UV experimentation, from a prof at the Rochester Institute of Technology.
- A tutorial piece at Photo.net.
- You'll probably also want to check out pages at NatureBlink, Hank Hogan's pages on Ultraviolet Nature Photography, and some nice photos by Fumio Yozokawa. WJ's Photo Homepage has a few bits about UV, and much more about infrared.
- Info about the legendary 105mm UV-Nikkor quartz lens.
- A page on UV lens alternatives.
