The basic building blocks of a digital image are the pixels, which in turn are made up of colour channels, or the primary colours of Red, Green and Blue (RGB).
Clipped highlights and shadows
FIGURE 1 The Histogram palette can show the difference between a damaged and a correct image, even if we can't see this in the image ourselves.
The colour channels can be viewed with either the Histogram or Channels palettes. The Histogram palette shows the spread of pixel colour values over the tonal range in each of the separate prime colours. Ideally the Histograms for each channel would cover the full range from the black to white points and form an island shape. If the image has been badly corrected, with the shadows and highlights pushed into blocks of black and white, the Histogram will form peaks with a sudden cut off at each end, which is referred to as clipping.
NOTE: There are many exceptions to this basic rule, so take is simply as a rough guide.
FIGURE 2 If the channel Histogram shows that the tones do not cover the full range from the black to white points then the image is thought of as being flat. The image on the left has tones that cover only two thirds of the full range. The shadows are correct, but the highlights have no white values at all. The image on the right is the same, but the contrast has been improved so that the tones do cover the full range making it look brighter. However, not all the channels are equal as only the Blue channel goes from the black up to the white point. As the Red and Green channels fall short in the highlights the result is an image with a blue cast.
Some images should have channels showing flat detail. For example, a view of a landscape covered in fog will have poor whites and little in the way of dark shadows, if any.
Other images will naturally have unbalanced channels; sunsets views will have a pink cast, for example.
FIGURE 3 A common problem is a deliberate exaggeration of colour. This is often seen with blue skies where the tones used are values that do not exist in nature and can be very difficult to print. At first the problem may not be very noticeable, but looking through the separate channels soon reveals the error. In this example the Red channel has collapsed in the upper half of the sky (indicated), where the tones contain only green and blue values with no red at all.
FIGURE 4 In this example the Blue channel has almost completely disappeared so that in all the colours are made of just red and green values.
In some cases an entire channel can collapse resulting in extremely vivid colours. Whether this is visually acceptable or not is subjective, but do not expect the image to closely resemble the colours in the original subject.