Into the infrared
Mike Bedford shows you how to take infrared photos with an ordinary camera or phone, and process them to produce some startling effects.
Mike Bedford shows you how to take infrared photos with an ordinary camera and process them for startling effect.
Visible light – the part of the electromagnetic spectrum to which our eyes are sensitive – ranges from a wavelength of about 380nm to just beyond 700nm, with violet at the short wavelength end of this range and red at the opposite end. Because objects differ as to which wavelengths they reflect, we perceive them as having different colours. We also receive ultraviolet and infrared radiation from the sun, so it’s interesting to speculate how the world would look if our eyes could see radiation at these wavelengths.
The concept of completely new colours might be a tricky one to get your head around, but photography is able to capture images in these invisible parts of the spectrum, with some interesting effects. Ultraviolet photography is difficult to achieve because glass lenses absorb it, but infrared is another matter entirely, because most ordinary cameras are sensitive to infrared. Here we’ll look at the practicalities of taking infrared photos, before delving into how to process them to produce dramatic results.
Into the infrared
We don’t need to know a lot about the physics of infrared radiation to make use of it, but a few facts and figures help set the scene. Infrared – commonly abbreviated to IR – covers a large range of the electromagnetic spectrum from just beyond red light, at just over 700nm, through to 1mm. We are interested in only part of that range, the so-called near IR, which extends from around 700nm to 1,400nm.
Important features of near IR, which are key to infrared photography using an ordinary camera, are that it can be focussed by a camera lens, and it can be detected by the imaging sensor in a digital camera. In fact, because cameras would otherwise be so sensitive to IR, and this would manifest itself as an unnatural red cast to photos, cameras have something called a hot mirror filter in front of the sensor. This allows visible light to pass while blocking IR – but, fortunately for us in our quest for IR photography, commonly it doesn’t totally exclude IR. The presence of that filter does have some consequences, though, as we’ll see later.
So cameras are potentially able to see IR, but that isn’t all that’s necessary to take IR photos. Because of the hot mirror filter, cameras are much less sensitive to IR than to the visible spectrum, so visible light swamps it out. For this reason, to take an IR shot it’s necessary to add another filter over the camera’s lens, this time to block visible light while allowing IR to pass through. See Infrared Pass Filters on the facing page for more.
Shooting in infrared
So we’ve seen the basic principle of IR photography – putting a visible-light-blocking filter over the lens of an ordinary camera – but the practicalities are a bit more involved. First of all, while most cameras and phones are sensitive to IR, it would be good to check that your
own camera can definitely see it before shelling out on a filter – even though it might only cost a few pounds. To do this, take your camera and a TV remote control unit into a darkened room, point your camera at the end of the remote control, and press any button on the remote while observing the camera’s LCD screen. If your camera is sensitive to IR, you’ll see a bright spot of light. If you have a DSLR without Live View, you’ll have to take a photo to look for that spot of light. From here on we’ll assume you’ve got a suitable camera and an Ir-pass filter.
Next, we have to consider exposures. Because of the internal hot-mirror filter, ordinary cameras are not nearly as sensitive to IR as they are to visible light. This means that you’ll need to use either a higher ISO rating, a larger aperture, a longer exposure or some combination of these. Remember that a high ISO will introduce noise into your photos and a large aperture will decrease the depth of field, something that is best avoided as we’ll see when we look at focusing.
In many cases, therefore, your best option will be to use a long exposure, which might be as much as a few seconds, even on a bright sunny day. In this case, you’ll need to use a tripod, but these don’t have to be large and expensive affairs – you can pick up small tripods for just a couple of pounds, and some can even be used with a phone. Note that if you use the default settings on most phones and many cameras, it will automatically select a high ISO rating under these conditions, so you’ll have to get to grips with the manual settings to configure it in a suitable way for IR photography.
The other technical aspect of capturing IR photos that we need to give some thought to is focussing. In our experience it usually works out OK, but we do need to bear in mind that IR focuses differently from the visible light that lenses are designed for. How much of a problem this is will differ from camera to camera, so a bit of experimentation is probably called for. If you struggle with automatic focussing, you might have to delve into manual focussing, if your camera has it. One useful tip, irrespective of your camera, is to use a small aperture (large f-number), because this will improve the depth of field, thereby eliminating problems of slight focussing errors.
With the technicalities out of the way, a few words on appropriate photographic subjects are appropriate. As you’ll see from the black and white image (page 60
bottom left), the two most characteristic features of IR photos are blue skies that are rendered very dark or almost black, and foliage that’s so bright that it appears to glow, especially in bright sunlight. So, while not completely excluding other seasons, spring and summer are undoubtedly the best times of the year because of the leaves on the trees, and the greater likelihood of blue skies and sunshine. Because of the possibility of long exposure times, though, avoid scenes with movement, even drifting clouds in the sky.
Finally, we should point out that when we look at processing IR photos, some of the techniques we’ll describe involve combining an IR photo with an ordinary photo of the same scene in various ways. If you want this option, you’ll need to take a pair of photos, one with the IR pass filter in place to capture the IR, and another without to capture the visible light. Needless to say, you need to be careful not to move the camera while removing the filter – but also remember that this takes time, so any movement in the scene is more of a problem. Even slowly moving clouds might not line up correctly between the two shots. In addition, use the same aperture for the two shots, because it affects the depth of field.
Basic processing
Most photos are not at their best straight out of the camera, and IR photos are no exception. In fact it’s probably true to say that IR images benefit more from a bit of post processing than visible-light photos. Here we’re going to start by explaining how IR monochrome photos can be made even more dramatic by image editing, before going on to look at some more advanced techniques involving Ir/visible pairs of photos.
Our package of choice is GIMP, which is highly respected – although you should be able to do what we describe using most fully featured photo manipulation software. Even entry-level software will be capable of working with a black and white IR photo, although the decompose and compose functions of GIMP – which
are required for working with Ir/visible pairs – probably won’t be available in more basic photo editors (or may be called something different).
A basic IR photo is a black and white image and while this might be considered the poor relation to a colour photo, black and white photography can produce very impressive results. However, because of the lack of colour information, contrast is much more important. Due to its white clouds against dark skies, and the ethereal glow of foliage, IR photography is inherently high in contrast, but even so a helping hand can often be used to good effect. In a nutshell, that is the essence of IR monochrome processing. As with all photo-editing tasks, remember to make a copy of your photo, and work on that rather than the original. This way you can try and save several different versions, but you’ll always have the option of reverting to the original.
First of all, although we’ve described an IR photo as black and white, the term ‘monochrome’ is more appropriate. Although a few cameras produce IR shots that look essentially black and white, in many cases they will be in shades of red, orange or magenta. Your first job, therefore, is to convert it to a greyscale image. Be sure to do this even if the image looks black and white, because it will invariably have some slightly coloured tint. This is achieved in GIMP by selecting Image > Mode > Greyscale.
Next it’s time to make the whites whiter and the blacks blacker, something that will normally make a good IR shot even better. There are several ways to do this but be careful not to go too far because this will cause detail to be lost in the lightest and darkest areas – though you might decide that losing a bit of detail in the sky is a price worth paying to make it appear really dark. The easiest way is to increase the contrast at Colours > Brightness-contrast… . However, for more control, go to Colours > Curves… and edit the curve from its default diagonal straight line. The best way to learn how this works is to try it out yourself but, generally speaking, the end result will be an approximately S-shaped curve.
Colour options
In the days of IR film photography, Kodak produced a colour slide film called Ektachrome Infrared as an alternative to black and white IR film. This created a false-colour image in which IR appeared as red, red visible light appeared as green, and green visible light appeared as blue. Usually a yellow filter was used over the lens to exclude the blue light. Commonly this was used for scientific purposes – maybe for detecting disease in crops by aerial photography – although limited use was made of this film for artistic purposes, for example the cover of Frank Zappa’s Hot Rats album.
If you take a pair of photos of the scene, one in IR and one in visible light, it’s fairly straightforward to process them to simulate Ektachrome Infrared. You might find the result a bit gaudy but it’s interesting nevertheless, and it provides experience of using
GIMP’S decompose and compose functions that you’ll need for some of our other more advanced effects. Before starting, though, it’s important to recognise that the camera might have moved slightly between taking the two shots. If so, you’ll need to edit one of them to achieve accurate registration.
The process involves extracting the red and green information from the visible light photo and combining these, plus the IR photo, to create a full-colour photo with the channels swapped to match the characteristics of Ektachrome Infrared. Now let’s see how to do that in
GIMP. First open the visible image. Next, select Colours > Components > Decompose… (in some other packages this function is called ‘split channels’ or similar). In the Decompose window, ensure that RGB is selected as the colour model and that ‘Decompose to layers’ is ticked, before clicking OK.
You’ll see that three monochrome images are generated as layers, containing the red, green and blue components of the original colour image. Next, open the IR image as a separate layer, using File > Open as Layers… . Note that this must have been converted to a monochrome image, and you might also want to enhance the contrast first, as described previously.
Now select Colours > Components > Compose… (in some other packages this function is called something like ‘combine channels’) and then, in the Compose window, ensure that RGB is selected as the colour model. Finally, select the IR photo against Red,
the decomposed red layer against Green, and the decomposed green later against Blue, before clicking OK. Your Ektachrome Infrared-like image will appear.
You might like the look of the Ektachrome Infrared effect despite its garish nature, but there are other ways of combining IR and visible images that you will probably find more aesthetically pleasing. The first of these methods involves keeping all the colour information of the visible image, but substituting the lightness information from the IR photo to get IR’S characteristic light foliage and dark skies.
The general procedure is similar to that we employed for creating the Ektachrome Infrared effect, but with an important difference. Instead of decomposing into RGB layers, we decompose into HSV layers (which might be called HSB in other packages), and similarly we compose using the HSV model. HSV stands for Hue, Saturation, Value (or Brightness in HSB). Hue is what we generally think of as colour, saturation represents how strong the colour is, and value is how light it is.
Having decomposed to H, S and V layers, the image is composed, keeping the hue layer for the hue and the saturation layer for the saturation, but using the IR image for the value. We should also point out that you can decompose to HSL (Hue, Saturation, Lightness) which, despite the similar name, is a different colour model to HSV/HSB. Now, of course, you’d replace the lightness of the visible image with the IR photo and, in so doing, achieve a different effect.
Getting more creative
If you look at IR photos online you’ll find lots of falsecolour images which share much of the appeal of typical IR monochrome photographs, such as apparently glowing trees. Usually, skies are blue and foliage is a very pale orange hue, although this can differ depending on how it’s been processed. Unlike the examples we’ve seen so far, these are not usually the result of combining IR and visible shots, but have been created purely by processing an IR image. Having said that, remember that the R72 IR pass filter allows a small amount of deep red light to be captured along with the IR. Some advocates of false-colour IR use a filter that passes even more red light using, for example, a 655nm filter. The reason that it’s possible to create a colour image this way is that the red, green and blue elements of the Bayer filter, which is used to separate data for the three primary colours in a digital cameras, have secondary peaks in the infrared spectrum.
As a first step, we suggest that you try different ways of mapping the IR and colour channels into the visible spectrum. You could think of this as a modified Ektrachrome Infrared approach. From an artistic viewpoint, the way in which grass and trees appear as magenta is probably the main objection to Ektachrome Infrared. However, at least one different mapping produces an effect which, in our opinion, is more pleasing. If IR is mapped onto red – the same as with Ektachrome Infrared – but green maps onto green and blue maps onto blue, foliage now comes out as orange, while the sky is still blue. Not only is this a potentially pleasing effect in itself, it’s also a first step into emulating the false-colour IR effect.
Next, to reduce the strength of the bright orange of the grass and trees, the saturation of the image can be reduced at Colours > Hue-saturation… . In the Huesaturation window, adjust the Saturation slider down from its initial middle value. If this reduces the saturation of the blue sky too much, you could try selecting only some of the primary colours to adjust, before decreasing the saturation. In most cases you probably won’t get quite the ethereal nature of light foliage that false-colour IR photos achieve, but there are lots of other things you could try, limited only by your imagination. One such possibility is to decompose the image again to HSV, after creating a modified IR image (but probably not desaturating it), and then compose with the IR image substituted for value.
With ways of digitally manipulating ordinary photos limited only by the imagination, the appeal of bizarrelooking images can easily fade. However, infrared photography – involving as it does, the capture of real images which can’t be seen with the naked eye – still has considerable appeal to many. We trust that this introduction will inspire to try it out yourself and, remember, this guide only provides some suggestions on how you could process your creations. Here the sky is most definitely the limit.