Getting the range
First off I need to explain something about dynamic range. A camera is a relatively simple being. It can only see a certain difference in light levels, its dynamic range, in other words it cannot see bright white to deep black in one go. Take a look at this which shows the problem:
We can either have the room properly exposed or outside properly exposed but not both with one picture. To our eyes we can see both room and out the window (although out the window may be bright enough to make our eyes hurt); this is because our eyes have a higher dynamic range than our cameras. In other words our eyes can see almost all the way from dark to light. To use my word analogy, from my intro to post article, a human would see: 'Hello World' where as a camera may see 'hll wrld' when under exposed or 'eo o' when over exposed.
HDR or High Dynamic Range imagery is a trick to get all the different light levels in one picture (e.g. to display 'Hello world' from one photo that says 'Hll wrld' and another that says 'eo o'). Normally it has to be done in post although some cameras (iPhones for example) can do it in the camera. I have never been that impressed with the HDR function on my iPhone however.
There are several different packages out there that simplify the HDR process and allow you to do funky things with the images but I am going to show you a basic (for an advanced post processor of images) way to make an HDR photo.
See the light
First step is to get our photos. To do this you can either use your camera's Auto Exposure Bracketing (AEB) mode or take three differently exposed photos yourself. This method needs one under exposed image, one mid range image and one over exposed image. (One point to note is that I too badly over exposed my over exposed image which causes us some issues but this can be ignored for this tutorial, the method is what matters!) When taking these pictures it is vital that you don't move the camera between shots as there is no easy way to line up the pictures laterally in the GIMP (at least not that I'm aware of) and forward and backwards motion is impossible to rectify. Not having the pictures lined up properly can cause horrible ghosting effects around the edge of objects and can result in a loss of definition. Also objects that move during the exposures will cause similar problems (e.g. plants, people, animals etc.) so it's always worth trying to avoid things like this when a blurred object will spoil the image.
Once we have the images we need to import them into the GIMP as separate layers. To do this open one image in the GIMP then go 'File > Open as layers...'. Select your other two images and press 'ok'. I normally open the mid-range image first, rename the layer as mid then open the other two images as layers, renaming them as dark and light or similar.
Once you've done that you should see something like this:
Now this looks nothing like what our eyes see, we still have some work to do but you can clearly see the three layers in the layer toolbox and a messy image in the image window. So what next?
Now we have to apply some masks to the layers. Masks allow you to selectively modify the transparency of certain areas of an image, in the case of HDR images by their whiteness. First up is the dark layer, this is relatively easy!
Make sure the dark layer is the one that you are working on (i.e. it is highlighted) then go to 'Layer > Mask > Add Layer Mask...':
This brings up the following dialogue box:
Select the 'Grayscale copy of layer' (American spelling again!) but make sure the 'Invert mask' box is unticked like above, and press the 'Add' button. This will give you an interesting result!
Looks like an old fashioned negative right? Also note how there are now two thumbnails shown for the dark layer, the one on the left is the layer and to its right is the mask. We're about halfway there now. We need to do a similar thing for the light layer.
Select the light layer so that it is your working layer and again go to 'Layer > Mask > Add Layer Mask...' and select 'Grayscale copy of layer' but this time make sure you tick the 'Invert mask' box like shown below and then press 'Add'.
This gives you our almost finished HDR shot (note how both the light and dark layers have their masks shown next to their layer icons and note the difference between them):
The final thing to do now to complete the HDR (ignoring any re-sizing or watermarking) is to flatten the image to allow us to save in a sensible format. To do that go to 'Image > Flatten Image' and the three layers will be compressed to give our final image with one layer:
Now I mentioned I made a mistake with over exposing too much, what this has done is caused the right hand side of the window reveal to look very grey instead of a bright magnolia and the same has happened to the boxes on the left hand side of the foreground. With a bit of patience I could have done better with my exposures but what you can hopefully see is that now, in one image, we can see the detail both inside the room as well as outside the window with no appreciable loss of sharpness or anything else.
HDR is a very powerful tool and if time and energy allows I may do some future blog posts on using some of the other tools out there that make this process easier and or examine some of the other ways to make HDR more punchy. A word of warning however, HDR pictures are large, very large, this is because they have the information of three images in one, used sparingly it can be effective, used a lot will fill your hard drive!
There seems ti sometimes be too many HDRs flying around the place and it can become a bit much but each to their own! Have fun playing with this technique and happy snapping!





















