Understanding Auto Focusing Lenses

Article Source: Best Digital Camera Discounts. Copyright 2008
Date: September 3, 2010

Virtually all but the cheapest digital cameras have some form of auto focusing feature built into them. In practice, using the system is fairly easy and intuitive - we are going to let the camera decide where the object is that we are photographing, so that it can then adjust the focus accordingly to capture the sharpest possible image. But how does this actually work?

The explanation of how auto focusing works is actually tied to a concept that is getting more exposure in newer digital cameras: the histogram of pixel intensities for a given digital image. You can view this histogram on your LCD screen after you have captured an image. What it shows is a graphical display of the number of pixels that have recorded a given brightness value in the image.

Basically the graphical display is a series of thin vertical bars stacked side by side. On the leftmost side of the display is the bar showing the number of pixels that recorded a dark value (no light was captured by these pixels), while on the very right of the display is the bar showing the number of pixels that recorded the brightest possible value. As you move from left to right, the intensity associated with the pixels increases, and the height of the bar indicates the number of pixels which recorded that intensity.

So what use is viewing a histogram?

Well, if an image is excessively underexposed, virtually all the pixels will be dark and the vertical bars in the histogram will all be pushed to the left. Alternatively, a histogram with all the bars pushed to the right suggests that most of the pixels recorded a high light intensity and therefore the image is probably over exposed. Most properly-exposed images show a distribution of pixel light intensities that are crowded toward the middle of the histogram.

There are exceptions to this rule. For example, if you shoot a picture of a model silhouetted against a bright window, most of your pixels will either be on the underexposed model, or on the overexposed background provided by the window pane. The histogram will therefore show a sizable number of bars at both the left and right of the histogram, and nothing much in its center.

For regular photographic scenes, however, the histogram offers a great way to size up the overall exposure of the image in a non subjective way. Moreover, the histogram provides a quantified measure of image exposure that the brains of the digital camera can use to understand what it is that it is looking at.

This insight into the nature of light intensity histograms is the key to understanding how auto focusing works.

Some of the newest high-end digital SLR models feature over fifty separate auto focusing areas in the metered image. This means that most of the image can be metered before a final focusing length is selected for the shot. For purposes of this article we need only consider how one of these auto focusing, or AF, elements does its job.

The metered area of a single AF element might represent only one percent of the overall image, but it nonetheless represents a tiny digital image in itself and it has its own little histogram associated with it. So how might looking at a histogram tell us whether or not the image it represents is focused or not?

To make the explanation as simple as possible, let's suppose the AF element contains a dark insect, a fly, hanging in the air far from the background elements behind it which have merged into a light blurred backdrop. When the fly is in focus, the contrast between the fly and its background is very distinct. There are dark pixels (the fly) and light pixels (the background). Sound familiar? The histogram would have peaks in the lower and upper parts of its intensity distribution, similar to the silhouetted model standing in front of the window pane.

Now consider what happens as the AF element containing the fly is defocused. The image grows increasingly less distinct as the fly blurs into a grey smear that now diffuses throughout the entire element. The corresponding histogram shifts from a bimodal distribution to a much more uniform one as the peaks spread out toward the center of the histogram.

This is how the camera can quantify the degree of sharpness, or contrast, in any given image element. Auto focusing works by optimizing the histogram for non uniformity in the distribution, which it assumes defines the best focusing.

In practice the method works fairly well. But there are situations in which the surveyed scene does not lend itself well to accurate auto focusing.

For example, poorly-lit scenes offer low contrast, and therefore make the point of sharp focus hard to define. Other times, if multiple AF elements are being used to select the best focus, the background might provide the most contrast in the image and the object you are hoping will be targeted for focusing is actually ignored because its contribution to the auto focusing argument is outweighed by the contributions due to the background.

If you were trying to focus on a hummingbird that hung a few feet in front of a tree you might be surprised when the bird blurred into a non distinct spot and the leaves of the tree rendered perfectly sharp - the precise opposite of what you had hoped would happen. In cases like this you need to switch off the multiply metered AF elements and use a single AF element to monitor just the object you are intending to focus upon. You might use just the center AF element in a case like this.

Auto focusing is an extremely useful element of modern photography, virtually indispensable to practitioners of sports and wildlife photography. But it also proves handy in a wide range of everyday photographic situations, since the world around us is rarely ever at rest. So learn to appreciate the auto focusing mechanism in your camera, understand its strengths and limitations, and in the long run this will help you to increase the quality of your photographic portfolio.