The histogram (left) in digital photography is a column graph that displays the distribution of dark and light tones of in an image. The left hand side of the histogram shows the darker tones with 0 equaling black, meanwhile the right-hand side of the graph represents the highlights in an image with 255 equaling pure white. The centre of the histogram shows the mid-tones (shades of grey) of an 8 bit image. The height of each bar represents the brightness of each tone (luminosity).
How these bars are distributed produces a distinctive shaped graph for a variety of exposures. A photographer can check the histogram on the camera to determine whether the exposure is correct.If an image has pixels which are packed towards the left-hand side of the graph then the image will be under-exposed with no highlights and little detail in the shadows. This occurs when there is insufficient light reaching the camera’s sensor. Conversely, if the pixels are squashed up on the right side of the histogram then the photograph will be over-exposed, with too much light reaching the camera’s sensor. Therefore there will be no dark shadows and the highlights will be blown out.
The video clip below, by Jean Francios O’Kane provides a brief introduction into reading a histogram. O’Kane uses the following analogy to explain over and under exposure. If pixels were stones and the camera was a shovel then a good exposure would occur when the shovel is able to pick up all of the stones, i.e. a camera’s sensor reading all of the pixels. However if the stones are arranged in a way that when a shovel is used it misses the ones on either side, then this is the equivalent of pixels being lost from an image.
A high contrast exposure would produce pixels that are spread right across the histogram up to the edges of the graph. Examples of this type of image could include (Freeman 2004):
- a large bright subject with a dark background
- a large edge-lit subject with a dark background
- a large dark subject with a light background
- a small bright subject with a dark background
- a small edge-lit subject with a dark background
- a small dark subject with a light background
Freeman (2004, p. 34) warns of the danger of photographing high contrasting situations because there is a:
risk of losing detail because their brightness range is likely to be beyond the dynamic range of the camera’s sensor.
This image illustrates the ‘hot’ (over-exposed) areas in red and ‘cold’ (under-exposed) areas in blue. The bright sunshine reflecting off the sea water and the pale tones of the people has created a peak at the highlight (right-side) of the histogram, with highlights clipped. The sunlight has also created a large shadowed area underneath the platform. This is illustrated by a peak on the left-hand side of the histogram, where shadows have been clipped. The darker tones (blue) represents a larger proportion of the image, showing a gradual decrease in pixels from black to mid-grey. Towards the highlighted end of the histogram the curve becomes steeper.
By reducing the exposure by 1 stop the shaded area has not been altered too much. However the highlights have been darkened, especially in the water. This is shown in the histogram as a ‘bump’ before dipping towards the edge of the right-hand side of the graph.
Alternatively, by increasing the exposure by 1 stop, the shadowed region underneath the platform has almost been eliminated. This has altered the histogram by making it rise up from 0 rather than starting from a large peak. The greys are almost consistent throughout the histogram before tailing off, with the platform clearly in view in the image. The increase in exposure has brightened the water that is lit by the sun and the tones of the people. This is represented by a sharp peak at the highlights end of the histogram, with highlights clipped.
An average exposure has vertical bars from lights to darks along the histogram, with the centre peaks towards the middle of the graph.
This is an average exposure with no extreme highlights or shadows. The water, umbrella and sky are represented on the lighter end of the scale, whilst underneath the fountain and the doorways in the background are darker, but only a small proportion are black, because few pixels are shown on the histogram at 0.
Reducing the exposure of the image by 1 stop has pushed the vertical bars of the histogram to the left, increasing the peak of the darker tones, particularly underneath the fountain and the shadows on clothing. This under-exposure has moved the lighter tones towards the mid-tones, with a sharper peak. This is largely due to the sky taking up the top third of the image, which now looks greyer than in the first photograph.
Lightening the image by 1 stop has had a dramatic change on the photograph and it’s histogram. The whole sky has been flagged up as ‘hot’, along with the water and the umbrella. This is illustrated by pixels being pushed towards the lighter highlights of the histogram, with a peak at the very edge which extends above the height of the graph, clipping the highlights. The increase in exposure has reduced the shadowed area underneath the fountain which has created a gap on the left-side of the histogram. The mid-tones are now more evenly spread as the shadows have been lightened, revealing details in the lower third of the image.
These types of exposure are flatter than average, with more values squeezed together in the histogram.
Finding a low contrast image was difficult at first, however I eventually came up with this image of a seagull against a blue sky.
The grey of the seagull against the blue, cloudless sky has created a histogram with pixels packed together in the centre. There are no black or pure white tones, so there is a large gap at either side of the peak.
Reducing the exposure by 1 stop has produced a straighter section on the histogram, with an almost even tone across the sky. The gap from 0 is considerably less than in the image above, but there are still black tones. The side-lit edge of the seagull has created a light grey tone which is a small proportion of the photograph.
An increase in exposure by 1 stop has moved the pixel bars to wards the right. The shadows have been lightened, which is shown by a gradual incline on the left of the peak and revealing more detail on the bird’s wings. The grey tone of the seagull and the consistent tone of the sky has produced a narrow band in the mid-tone region of the histogram, with no white visible.
This exercise has enabled me to assess when my images have been either under-exposed (pixels pushed to the left of the histogram) or over-exposed (pixels pushed to the right of the histogram) at the point of shooting. At this point in my workflow, interpreting the histogram correctly will help me determine which images to keep and which need to be deleted to free up memory. Using the highlight and shadow clipping warnings during the image editing process will help me process my images correctly.