A Tutorial for Digital Infrared Photography

Tübingen Beyond Red

In this brief tutorial you will learn the basic steps from the image out of the camera to this final with a couple of mouse clics using free software only.

Out of the camera infrared image


You will have to start with some special hardware. An infrared image requires an infrared filter mounted to the camera. The issue with modern digital cameras is, that their built-in AA filter will block almost all infrared light. If you now add an infrared filter to the lens, that blocks everything but infrared light, then there is not much left to reach the sensor. This works in principle but the poor transmission rate of both filtes in a row will result in very long exposure times.


infrared image out of the camera

With those long exposure times between 5s and 30s or even more, a good tripod and patience is required. In addition to that a human eye cannot look through the view finder when the infrared filter is mounted. That means the image must be composed without filter, then the filter is mounted and after that the image can be taken. 

Naturally you cannot take this tripot into a small and moving boat nor shoot moving targets. Here a converted Canon EOS camera comes to the rescue. In such a camera the build-in AA filter was exchanged in favour of a now built-in infrared filter. I choose a 700nm filter but would go for 680nm next time. With such a camera the exposure times are quite normal and the photographer can actually look through the view finder again. 


This reddish image is hardly usable. An extreme white balance makes all the difference. Just to be save I shoot thes pictures in RAW format which allows me to set the white point later at my pc. There I try spots within the leaves as white point. For the RAW conversion along with the white balance I usually use the free converter from canon, Digital Photo Professional. There are more free raw converters available. Actually the open source community makes a good job in delivering programs like UFRaw or RawTherapy. But not all of them are suitable alike for those exreme white balances as they are required for infrareds. The result is like this: 

White balanced in dpp

Next you go to the color mixer and switch the red with the blue channels, resulting in ... 

Red and blue channels switched

At this stage an infrared image is usually a bit soft, that is why I improved the contrast. 

Contrast enhanced

First I thought that is about it. Sometimes I add some vignetting here. 

Well, this is not any vignette, that is my special vignette. It only darkens the shadonws but does not darken the overall image. To achieve this I do the following: 

  1. Copy the image into a new layer (new from visible).
  2. On the new layer reduce the saturation to zero.
  3. Set the layer mode from "normal" to "mutliply"
  4. Don't get shocked here.
  5. Add a layer mask.
  6. Fill the layer mask with a circular gradient from black in the middle to white at the edges.

Vignette added

A vignette like this is nearly invisbile, although it leads the eye to the center of the gradient, the middle of the image in this case. I had a look at it and felt, that in this case it adds some depth to the image. I wondered how more of that vignett would look like ...

Add one more vignette

Now I had the Sisters of Mercy in mind ... I NEED MORE!

I need more!

 This is about right but a bit dark in total. So I used the gradation curve to lighten the image and erased a contrail from the skye and it was finished: 



The Channel Mixer

Just to give you an  understanding on how the channel mixer can influcence the image see the following example.

Different results from different cannel mixer settings

To the left you see the original image which is a typical result from the white balancing.

The second one is the standard usage, red and blue channels are switched.

The third one has red and green mixed.

Blue and green are mixed in image number 4.

The next one shows a mix as follows: Red -> Green, Green -> Blue, Blue -> Red

The last one is Red -> Bllue, Blue -> Green, Green -> Red