2. HDR for grown-ups
Treat HDR as it was intended – a way to handle a high dark-to-light range
HDR was invented to archive light, but got hijacked and corrupted. Here we return to what it’s really all about – skilfully creating images that span the range of light without getting over-excited. The HDR part of the process is dead simple: you shoot a range of identical frames with different exposures, sufficient to capture everything, and then one-click software turns them into a single image file.
What many people and manufacturers call HDR style isn’t that at all. The style of revealing every tone in the scene (and then some) is down to tone-mapping, which is the set of algorithms for displaying the image in a viewable way on a screen or in a print. There’s a hard-to-resolve issue here: you have an image file that contains everything, but your screen or printing paper can show just a fraction of it. Your eyes, on the other hand, which work quite differently from a sensor, can take in any amount of dynamic range. So, there has to be a compromise, and in professional work that usually means taking traditional photographic values as a base – and not straying too far from them. The argument is that after a couple of centuries we’re happy with a photographic ‘look’, even if it includes flare in the highlights and dense blacks. The aim with realistic HDR shooting is to stay within those viewing borders of tolerance.
There are many ways of doing HDR, but the following (shown below) is at least one solid and restrained way to produce a normal-looking image. The key is to create a 32-bit floating point
TIFF file at the first stage, when you combine the different exposures, in Photoshop or other software. Then you continue the processing in Adobe Camera Raw (ACR, the normal RAW processing engine). The experience at that point is notably different, and the example here shows the steps. Normally you would shoot a range of exposures two ƒ-stops apart, that go from dark enough to hold even the brightest highlight, to bright enough so that even the darkest shadows are rendered as mid-tones. In this case, a Thai interior shot entirely in available light on a very bright day, I varied the shots by 1½ ƒ stops, and needed five frames to archive all of the light.
This method starts with making a base overall adjustment to gather the brightest and darkest tones, and because I’m dealing with what you could think of as a super-sensor, I do what I would never consider normally: take the Highlights and Shadows almost to extremes, then raise the Contrast. Ordinarily, this would produce the kind of hyperactive result I want to avoid, but with a 32-bit file it stays realistic. The real work is then done with radial filters. Here we need restraint. You can pull out anything from this file, so it’s essential to keep in mind what you want to get from the image, and go no further.