False-Color Infrared Imaging

I sat down a few days ago to write a lecture on rigid gels for Camberwell and a few hours later I was plotting to turn my old Nikon D40 into an infrared camera for false-color infrared (FCIR) analysis. Procrastination isn't usually this productive. FCIR is a nice trick that can be used for pigment identification, or to see areas of retouching/restoration in past, or distinguish between carbon black and iron gall ink, or probably some other things appealing to the book conservator that I haven't thought of yet. Some of these materials absorb IR light and some reflect it, making them appear black or light respectively, though your human eye can't tell. My new clever camera will be able to make a photo using these longer wavelengths, though. Once you have the normal visible light photo and the IR photo, you can combine them in Photoshop (this was a revelation) by mixing up the channels to create an FCIR image: from the color image, the green light is registered as blue, the red light is registered as green; and the information from the IR image is registered as red. 

Normally this is done with a specialized camera and software to go with it, but I discovered that you can do it with a normal SLR with the help of this page and worked out these options with the help of some old photography friends:

  • Use IR film: not convenient

  • Modify the camera internally to remove the bit that normally filters out IR wavelengths (not cheap)

  • Get a £34 filter that screws on the end of the lens and turns your world red. Bingo!

The downside of the filter is that it blocks a lot of the visible wavelengths so you can see hardly anything through it unless you point it straight at the sun. It also means very little light gets to the sensor, because the internal IR-blocking filter hasn't been removed, so exposure times are long. Because I'd be taking a normal photo first then screwing on the filter without moving anything, and I'd need the tripod anyway, I decided a little lack of visibility and extra exposure time was worth the approximately $400 and longer wait saved not modifying the camera. So I bought the filter (Hoya R72), got Saturday delivery thanks to that Amazon Prime membership I forgot to cancel, and just around sunset a little envelope fell through the letterbox. After a few red photos of my garden and the cat I got this:

One of my 18th c. French books, this one with a less usual marbling pattern (the colors are still standard issue). And yes, that's marbling on marble. 

One of my 18th c. French books, this one with a less usual marbling pattern (the colors are still standard issue). And yes, that's marbling on marble. 

The red is the result of the camera, built for registering color, getting to grips with the infrared. It gets turned into black and white in Photoshop first:

Black and white IR photo

Black and white IR photo

In case you don't believe your eyes, this (below) is what a black and white image is like taken without the IR filter and (below that) the normal color photo. 

Normal visible light black & white photo

Normal visible light black & white photo

FCIR - marbling - VIS.jpg

After I worked out how to split and recombine channels, I got a false-color image that pretty much means nothing without some reference samples but it nevertheless exciting enough to present with a flourish in class on Wednesday. I can see a lot of pigment grinding in my future. We can make a few observations now, though: the red has stayed red, the green has gone an indescribable shade of muddy yellow/brown/green, the yellow is a greener/bluer yellow, and (wait for it) the blue is pink. Ultramarine behaves like this, by the way, but I'd want to confirm with PLM/FTIR/XRF/more-than-two-hours'-experience-with-FCIR before making any wild claims. 

As I'm imagining how this is going to go down in class (me: why aren't you guys more excited??) I'm reminded of the ongoing conversation, at Camberwell and West Dean and in our profession as a whole, about how much science is enough. Because most of us came from arts backgrounds, the heavy science is generally a struggle, and since two years isn't enough for any of it the time that could be spent on science always has other claims to it. I think you just need enough to look up the bits you don't know—there will always be bits you don't know. Enough that you can know almost nothing about false-color infrared photography, and two days later, have a new toy.