First Impressions: Infrared by D90

Today was a beautiful, sunny day, just perfect to capture some infrared rays. So, outfitted with a Hoya R72 filter, I grabbed my D90 and decided to try my luck with some hand-held infrared photography.

I have had some experience with infrared every since reading a book and discovering that my P5100 could capture infrared. Having passed the “infrared remote” test, I decided to invest in a 52mm R72 filter. The results I got were interesting, however limited by the P5100’s eight second maximum exposure. I would set the camera to manual, select the widest aperture, and set the exposure to 8 seconds. Depending on how much light I had, I could vary the ISO to improve the quality, but it was always pretty high. Then I would set my camera on a tripod, shoot, and hope for the best. I was able to capture in color or black and white, and the color images were a deep monochrome magenta.

With the D90 and some further reseach, I discovered I’ve been making more work for myself than is necessary. With a maximum metered exposure of 30 seconds, up to 30 minutes of bulb exposure, an ISO up to 6400, and much more control over white balance, many more things are possible.

Using the same R72 filter, only in a 67mm size to fit my 16-85 f/3.5-5.6, I am able to capture hand-held images at about a 1/15-1/25 second exposure at 3200 ISO. The D90 has no problems with either focusing or metering. Depending on how I set the white balance, once I dial in the appropriate exposure compensation, I get consistent exposures.

It’s the white balance that’s the key, and I have more control over it with the D90, than with the P5100, so I have more variables to explore and overcome.

It seems that setting the white balance manually to 2500°K produces less saturated magenta images with good shape and detail, and distinct color information. While most of the image appears monochromatic, green grass and shrubbery appears a pale, desaturated green. Apparently setting the white balance this low forces some of the visible spectrum to come through. I’m not sure if this is true IR or not, but it is no less interesting. The grass definitely has that “Infrared” look, only in color. Nothing in Photoshop seems to improve this effect, although adjusting the tint in Camera Raw seems to control the saturation of this green, making it look less Infrared-like.

However, setting the white balance to the other extreme, 10,000°K, produces saturated red-orange images, not unlike redscale. And, longer exposures produce even more interesting color effects with colors ranging from highlight to shadow of white, yellow, orange, red-range and black, respectively. These images translate well to black and white images when you adjust the red, yellow and sometimes magenta sliders in Photoshop’s “Black and White” adjustment layer, or the red, orange, yellow and sometimes magenta in Camera Raw’s “Convert to Grayscale”. I think this might be a more faithful example of Infrared, as all the camera’s settings are coaxing it away from anything not in the red spectrum.

Another approach is to do as the books suggest, and perform a custom white balance off green grass or shrubbery. This yields an effect somewhere in between the two, with grass having a slight magenta hue, while the sky goes more of a red orange. This translates well to black and white also. I can simulate these settings without sampling by setting the white balance to 4350°K and adjusting the tint to the magenta side.

But, this is all in color, and there really is no color in infrared, hence the -red part. The main object of shooting in color is to get some color information, so I can tweak the conversion to black and white, or do some false-color post processing. The other option is to shoot in black and white, and let the camera do the conversion, and this works quite well also. With this option, I can choose from a yellow, orange, red or green filter, yet another set of variables. I can say for sure that the red filter cuts down on exposure times, with the green filter producing darker images. But, which yields the best contrast requires more experimentation.

And this is all with one lens. Other lenses will undoubtedly produce different results. Perhaps no lens (pinhole) or even plastic lens (Lensbaby) will produce more variations. I have much work ahead of me. For now, my goal is to be able to predictably produce these two distinct styles of color infrared images, by establishing the appropriate camera settings, and come up with an optimal setting for monochrome infrared images.