Here's a bit of background. First the purpose of using full spectrum photography (or any detecting device for that matter) is to extend our perceptual abilties. The operative notion is that some phenomena may be detected because it is at the fringe of our senses and may in fact not be visible or detectable by some standard devices. This reflects the idea that perception is not only what our eyes and ears for example detect, but also how that information is processed. Hence the idea of using devices that see into the margins as it were.

We tried this back in the 1960s with IR photography which had its own problems. I may have been the first to use thermal imaging in the mid 1970s, which was another attempt to see past our physical limits (that one did not work btw). And now we have digital photography and the extension from UV through visual on into the near IR as 'full spectrum' photography. Again trying to see out into those nasty margins. Experience with all of these technologies has indicated to date that we probably will discover nothing but new artifacts. This is what happened before. We all thought we had something, but upon closer inspection and with a fuller understanding of the technologies and their quirks, we came to realize that there was nothing truly 'unexplained'. In fact some of the most disturbing photos have been taken under normal illumination conditions with very standard photographic equipment.

Now as to the questions. The nature of the detectors used in the cameras is such that they are sensitive in the IR region in what seems at first a non-uniform way. This has to do with the physics of the detectors. The important thing to note is that the IR (near IR) signal is much stronger than the near-UV signal. And this means you have to go through hoops to be absolutely sure you aren't really seeing IR leaks. The UV is so weak that it can be easily overcome by IR light or even visual light. And that could be one of the first sources of artifacts.

As to how one detects UV (or IR for that matter) things become a bit dicey there as well. CCD cameras behave one way while CMOS cameras behave slightly differently. Ironically, one of the better cameras for the purpose otherwise actually shows the near UV as a range of colors ranging from pink through a deep violet. In other words, you can't be assured that UV signals will all render as you might think they would intuitively. To compensate, you can try to filter out the UV, which is the primary cause of most of the confusion. That way, if you see a strange pinkish glow you know you are really looking at a piece of the UV spectrum and not a piece of the IR spectrum.

What no one tells you is that you need extremely high quality optics to do all of this. I feel for people who have tried to do this with cheap cameras. What is takes is a camera that will run you about $400 or so with proper modifications, and anotehr $1600 or so in the lens. This is because you need to use very high quality glass (quartz lenses are common) to avoid creaming the weak UV signal. The lens I use is virtually flat optically from about 100nm up through about 1200nm, which means the glass is not getting in the way opticallly. You have to use equally high quality filters for precisely the same reason; don't want to spend all that on a lens and then screw up by putting a standard glass IR filter in front of it.

The same problems maintain whether the camera is a UV-only or a Full Spectrum. That IR is still out there to get you if you aren't careful. Once you understand how your camera behaves, it is a good idea to do color separation in software. Be careful though, because, yet again, you can introduce some lovely artifacts in that process.

Do I believe FS will actually uncover anything? I have this gnawing suspicon the asnwer is 'no'. But my eternal optimism hopes that I am wrong.