The past couple of weeks has seen me in testing mode for a few different topics. The main one being my recent purchase of a used Tokina 12-24mm f/4 lens, that I wanted to fully check for functionality before the return period expired. And I’m very happy to find that everything works and the lens is now a permanent part of my photo gear.
The other testing I have been doing is with the IR photography capabilities of my cameras and lenses. For some time now I have had an interest in the look of IR photos and wanted to take another attempt at taking some myself. To this end I purchased a filter to fit my new Tokina and attached it all to my D2X - and was not too happy with the result. Guess I should have done a little on-line research first as I would have found that numerous people have flagged both the D2X and Tokina 12-24 as not recommended for IR photography.
With the D2X, Nikon did a very thorough job of designing/building it to minimize the effects of IR light on the image sensor. This is wonderful for “normal” photos and especially skin tones but not so good for my testing. It’s not that it can’t take IR photos, just that they will require much longer exposure times to produce anything usable, which in turn increases the effective noise levels on the images.
Meanwhile my older D100 seems pretty well suited to taking IR photos. My test shots with the Tokina lens were relatively free of flare in the image center - unless I had the sun in the frame, or close to the lens axis that it was able to shine into the lens. However, I did notice one anomaly between the Tokina and my Nikon 60mm Micro - after they’ve been post-processed, the images from the Tokina are predominately blue, while the Nikon retains more red to the overall image (this is after doing a blue/red channel swap; I’ll cover this later.
One of the biggest problems for IR photography is getting the exposure “right” - right being a generic term as this is subject to the personal taste and artistic intention of the photographer. Yes, it is certainly possible to go to extremes of over/under exposure and produce a bad photo, but there is a significant range in which the exposure can reside and still look good.
Back on the topic, one of the first things to do is pick (or set) a white balance. From my online research it seems that the common consensus is to take a custom WB reading off green grass lit by the sun; or use the WB picker tool in editing software and take a reading off the grass (or similar toned subject). This is all well and good but once I have the R72 filter in place, neither of my cameras will take a custom WB reading. So, I tried the WB picker tool - and found an interesting thing. Using Lightroom to do this the WB was consistently being set to 2000K and -36 tone. Now if you look at the WB tool in Lightroom’s develop mode, it only goes down to 2000K - this is as red as it can go and is actually within the range of “Near IR” on the Kelvin Color Scale, which using Wein’s law is calculated at approximately 3000K. This got me to thinking, why not just set my camera’s WB to their lowest and use that for the photos?
According to the camera manuals, the D100 can be set to a custom WB of 2,700K and the D2X can be set to 2,500K. Add in an exposure compensation of around +2 and this is a pretty good “set and forget” setting. If the images are still a little too far off once in Lightroom you can go ahead and move the WB setting to 2000K and -36 tone.
Beyond the actual physical conversion of the camera body so that it can shoot IR photos without an additional filter on the lens, this is as close as you can get to working in the “Near IR” spectrum - and far from working in the “IR” spectrum as it resides at approx. 300K which is far beyond the ability of a DSLR regardless of filters. Not that any of us can see anything remotely near these wavelengths anyway.
As I mentioned earlier I wanted to expand a little on my comment on channel swapping. You will need Photoshop, or some similar editing program that has channel adjustments to carry out this technique.
- Open the image in Photoshop
- Apply a Channels adjustment layer
- Select the Red Channel and set Blue to +100, Red to 0
- Select the Blue Channel and set Blue to 0, Red to +100
- Flatten image
- Carry out further editing or hit Save
The idea behind this technique is that it will retain “blue skies” in false color IR images. Not something to worry about if you convert all IR images to B/W, as a lot of people do. The only thing that I’ve found that seems odd with this technique is that I can take the same camera and settings to take a shot, swap lens and take the shot again and the false coloring will vary greatly in the red-blue spectrum. My Tokina 12-24mm seems to produce images with a blue color cast over the entire image, while my Nikon 60mm Micro produces more of a magenta color cast but it doesn’t tint other colors in the same way as the Tokina seems to.
In general I’ve been a little disappointed in how the Tokina 12-24mm handles IR photography, which as it’s pretty much tagged as a “do not use for IR” lens across the internet is not surprising. In fact, my old (slow) Nikon 18-55mm lens produces more pleasing IR images with no apparent hotspots. So much so I’ve decided to return the new 77mm Hoya R72 filter and will instead get something more useful for landscape/architectural photos - a good 4 stop or higher ND filter or a circular-polarizer; I’m leaning toward the ND personally. I’ll still be experimenting with IR photography, just that I’ll be doing that on my “spare” body and lens combo - D100 and 18-55mm. I’ll keep the D2X and Tokina 12-24mm for my main camera and swap out for the D100 when it’s IR time.
Now I don’t really have much in the way of IR photos to share as examples as I was aiming at testing rather than aesthetics - in other words, they suck! As I get more experienced with shooting IR photos I will post examples here for everyone to view, just that at this time I’d rather keep them to myself.
I will share a few “normal” photos from the same testing sessions though (in my next blog post) as examples of what the Tokina is capable of.