, , , , , , , , , ,

rgbhistogramAn RGB histogram

A photographic technique that has recently piqued my interest in that of ‘exposing to the right (ETTR)’.

Rather than attempt to rewrite the theory behind this in a way that would almost certainly be inferior to the original explanation, have a look at these two articles on the Luminous Landscape website – this one, and, in particular, this one.

I’d heard about this technique before, but a recent podcast by photographer Martin Bailey went into the topic in some detail. While Martin is a nature photographer who frequently encounters high contrast scenes and scenes involving snow, the part of the podcast that particularly interested me was that he recommends shooting to the right of the histogram and darkening down the image in post-processing, to achieve the same effect as if you’d exposed broadly to the middle of the histogram in the first place. Why? Because of the amount of data contained in the upper parts of the histogram. Thus an image that is ETTR and darkened down in Lightroom would contain more data and be of a higher quality than an image that was simply exposed to the middle of the histogram in the first place.

There doesn’t, however, seem to unanimous endorsement of this technique in the photographic community. Chris Marquardt, of the Tips from the Top Floor photography podcast, said this topic was a little complex for an episode of his show, but offered me this in an email:

This is an area where Martin and I don’t agree. I’m of the “expose correctly” school, mainly due to the fact that the way modern cameras are implemented, they mimic film behavior to be more pleasing to the eye and therefore tend to only be linear about +/- 3 stops from the 18% (grey card, zone V) or 50% (digital) exposure.​ Exposure above or below those points typically leaves the realm of linearity, resulting in software having to do a lot of guesswork when pulling right-leaning exposures back down. One reason for the loss of detail beyond +3 stops is that individual color channels are beginning to get saturated. The way Lightroom/Aperture/etc. “magically” save individual blown-out channels during highlight recovery is by borrowing detail from not-quite-yet blown-out channels and worst case by inventing data based on surrounding image areas. So for me it’s expose correctly for whatever is important in the image to grab the maximum possible detail. If the subject falls beyond +/- 3 stops, reel it into that range. This loss of detail is also why the proper placement for textured white and textured black on the histogram is never as far left or right as possible. The fact that the histogram on your camera isn’t actually based on the RAW image, but on an already processed JPG hidden inside the RAW, doesn’t make things easier…

He also provided some practical advice when I replied asking about any evidence showing the results either way with regard to ETTR!

So, the next step is a test to find out whether ETTR actually makes a substantive difference to image quality – stay tuned for a test on these pages soon.