Sunday, November 25, 2007

More fun with ultraviolet

uv403_uv0_z.jpg

uv403_door_z.jpg

Continuing with the theme of the previous post, here are a few more photos with the fancy new UV & IR filters I bought the other day. What I'm trying to do here is completely nail down the fact that my puny little digital camera can, in fact, see ultraviolet light. Various self-styled experts will tell you it's impossible, their opinions seemingly based on a couple of minutes worth of Googling and zero minutes of actual experimentation.

I took a few photos of more or less the same angle using various filter combinations. The first three were with a B+W 403 on the camera. This filter blocks visible light, and lets through UV and a bit of infrared. The second three were with a Hoya R72 on the camera, which lets infrared through, and blocks visible and ultraviolet. The goal is twofold: First, to show that yes, the camera does detect something with the 403 in place. And second, to show that what it's seeing cannot be just the IR component.

In addition to these two filters, I had a Hoya UV(0), which just blocks UV, and a half-opened window, which blocks even more -- but not all -- UV. I used both, because in the previous post I showed the half-opened window trick using a black light bulb as an improvised filter, and I wanted to demonstrate that the effect is visible with a real UV filter on the camera, and occurs with a known UV-blocking material just like it does with a regular window.

So the top photo is with the UV(0), and the second is with the window. As you can see, the window blocks substantially more UV, which isn't surprising since it's also quite a bit thicker. It's also nobody's idea of an optical-grade material. It should be apparent that the left and right hand sides of both photos are different, both in brightness and in color. The UV(0) and obviously the window look completely transparent in visible light, so the difference must be due to either ultraviolet or infrared.

[Oh, and before anyone accuses me of image manipulation: I did brighten the pics up a bit in GIMP using the Levels tool, since the originals were rather dark and it wasn't easy to tell what was going on. But no color tweaking, and certainly no lightening of individual regions within a photo. Everything's just as it appeared, just a few percent brighter overall.]

uv403_r87_z.jpg

The third photo shows the R72 infrared filter being handheld in front of the camera. I'm not sure what this proves, exactly, but I was curious how it would look. It's black to the naked eye (as is the 403), but I wondered whether there'd be enough IR making it through both filters to make it transparent, or at least non-black in these circumstances. Apparently not, at least not in this photo. Possibly the auto-exposure is to blame, so I'll probably need to take a couple of additional pics with one filter directly on top of the other. I'm not sure what that will prove, precisely, but it might be interesting.

r87_uv403_z.jpg

So photo #4 is the reverse situation, with the R72 on the camera, and the 403 handheld. Black again, just like last time. You might notice what look like a couple of spots of light on the filter -- I'm pretty sure those are reflections off the back, not light passing through it.

r87_uv0_z.jpg

R72 on the camera, UV(0) handheld. This is to see whether the color & brightness differential is visible with purely infrared light. Sure doesn't look like it, does it? Regarding the top two photos, I said that the difference had to be due to either ultraviolet or infrared, and now here's what you get with the UV component removed. So if the difference isn't infrared, and it isn't visible, it must be the ultraviolet. QED, eh?

r87_door_z.jpg

And the same experiment, this time with the half-open window. The non-window part might be just a shade brighter, indicating the window might be blocking a little IR too. But it's nothing like what's shown in photo #2, and there's no color variance that I can see here.

No comments :