Linear capt ure and ga mma CU RVES

Cameras don’t work like our eyes – that’s why it’s easier to recover details from shadows in RAW files

We treat the way we see as natural and straightforward, yet it’s anything but. Even our basic response to different levels of brightness is complex. We see twice the brightness, for example, as only a slight increase, and we are aware of changes in dark tones much more than in bright ones. In other words, the response of our eyes to light is non-linear. A sensor, however, is linear. Twice the brightness is recorded as twice the signal. The result is that a linear image capture, which is what your camera gets as soon as you take a shot, would look very dark. You never get to see this, because the camera automatically applies a correction to make it look normal. To match the way our eyes work, this correction is not an overall increase in brightness, but an increase that favours the dark tones. In the form of a graph it is a steep curve, as shown above, known as a 2.2 gamma correction. Seen as a tonal range from black to white, the correction spreads the tones out evenly. This is why a RAW file allows you to recover details in an image that might appear to have been lost by a less-than-perfect exposure, particularly in the shadows. Applying a stronger gamma curve will pull more visible detail out from those densely packed shadow areas. In other words, most of the information in a just-captured image is clumped in the dark areas, and needs to be pulled out.