In Camera: Using Histograms to Improve Exposure Photojournalism
Histogram of an image A graph representing 256 levels of brightness Height of bar represents number of pixels Pure black Left edge Value: 0 Pure white Right edge Value: 255
Proper exposure This image has rich blacks and a bit of clean, bright white. It has good distribution of everything in between. The graph can be broken into five regions: blacks, shadows, midtones, highlights and whites. This histogram comes from the Levels dialog box in Photoshop. This can be adjusted in the Levels dialog. The histogram can also be viewed in the Histogram window. These are both identical to the histogram shown in the camera. When the histogram is broken into 20% regions, from left to right: blacks, shadows, midtones, highlights and whites. Notice how the shadow portion of the graph is not very intense and the highlight region is higher, so more intense. Both the left (black) and right (white) edge of this histogram have spikes, which indicates that a condition called “clipping” has occurred in both the shadows (left edge spike) and the highlights (right edge spike.) Clipping results in the loss of detail in the shadows or highlights. It is particularly difficult to adjust an image that is badly clipped in the highlights. There is no detail recorded in that region of the image and regaining details in the uniform will be difficult. It is possible to recover a little more detail in the shadows by adjusting the exposure or through adjusting this image in Photoshop.
Proper exposure In Program mode, a camera’s full frame light meter generally exposes the frame to a blended, uniform 40% gray, such as in this image. The histogram is a representation of the tones in the image, with black represented as level 0 on the left and pure white as level 255 on the right. This histogram has a pretty even distribution from left to right. There is a slight spike on the right indicating some clipping in the highlights. If you wanted to recover details out of the florescent light fixtures, it would be difficult to do.
Overexposure In most cases, the camera’s light meter properly exposes an image. This photograph was purposely overexposed by +1. In instances where the meter is tricked by difficult lighting situations, it is useful to know how to read a histogram to adjust the exposure. Overexposed images (histogram heavy on the right) are too bright and highlights can be “blown out.” This image’s histogram has a distinct shift to the right, away from the pure blacks.
Tricky lighting In this typical classroom scene, the photographer needs to consider which part of the image should be properly exposed. Do we want the lesson on the overhead, the teacher or the students to be properly exposed?
Underexposure In most cases, the camera’s light meter properly exposes an image. In instances where the meter is tricked by difficult lighting situations, it is useful to know how to read a histogram to adjust the exposure. Underexposed images (histogram heavy on the left, virtually nothing in the upper half) have a lack of detail in the shadows. The significantly overexposed white portion of the projector screen is clipped. The spike at the extreme right edge of the histogram indicates that clipping has occurred in this image. Trying to adjust that portion of the image to recover any detail of what is on the overhead projector screen will be difficult or next to impossible.
Tricky lighting This is what the Levels dialog looks like in Photoshop. This is the identical histogram that your camera shows. Notice how black this histogram indicates the image is. Also notice the very narrow bar indicating pure white on the right edge of the histogram. Otherwise, there are virtually no pixels in the upper third of the brightness scale. The clipping in the shadows is not extreme in this image, but trying to gain any detail from the clipped highlights of the overhead screen will be next to impossible. In this case, the photographer and the person using it in a publication need to decide if this image is satisfactory to illustrate what is desired.
Backlit scene Notice the full-height bars on each end of the histogram. This shows that the exposure has “clipping” in the black and the white. With very little detail in the face of the subject, this camera exposed more for the outside scene than the person, although if you average the entire image out, it is a properly exposed 40%. Clipping has occurred at both ends of the histogram, so there will be problems trying to gain detail from shadows and highlights from this image. These histograms will always look “U”-like. Adjust your exposure to overexpose the darker regions to gain the desired detail in the face of the subject. Either lower the shutter speed, increase the aperture (smaller f-stop number) or increase the ISO.