The image sensors used in most digital cameras were originally developed for use in camcorders, so they are very small. Sensors are measured on the diagonal. The Charge Coupled Device (CCD) image sensors used in most digital cameras measure 1/1.8'', or .555'', or 13.7mm. For comparison, a 24 x 36mm (35mm) film frame measures 1.70'' or 43.2mm on the diagonal (see Figure 3.1).
Smaller image sensors have some distinct advantages, but they pose a number of design problems, as well. Smaller sensors enable manufacturers to sell the same sensor to both video and still camera makers, reducing costs through economies of scale. They also require smaller lenses, which reduces the size and weight of cameras. But the pixels on a small sensor are very close together, which in turn requires that the lenses used with small-sensor cameras be of very high quality. In fact, a big part of the cost of a 3-megapixel P&S camera is in the lens.
Figure 3.1 The CCD and CMOS image sensors used in most digital cameras are much smaller than a 35mm film frame.
Where'd My Pixels Go?
When comparing camera specs, you'll often see numbers for "number of sensor pixels" and "effective pixels." The latter number is the actual number of pixels produced by the camera, and it is always lower than the first number. So where did those pixels go?
Some of the pixels at the edge of the sensor have an opaque black dye that blocks light from hitting those pixels. This is done so that the camera has a reference point for the darkest part of the image.
Some cameras lose some effective pixels because the camera's lens can't cover the entire sensor area. Canon's PRO90IS, for example, uses a 3.3-megapixel sensor, but the lens-borrowed from one of Canon's DV camcorders-is too small to cover the entire sensor. As a result, the camera only uses the central part of the sensor, giving an effective pixel count of 2.6 megapixels.
The individual pixels on a small sensor have less surface area than the pixels on a larger sensor. As a result, they capture fewer photons, which makes them less sensitive to light than large image sensors. Many CCD sensors employ a grid of tiny lenses-one per pixel-called a microlens array (see Figure 3.2). The microlenses are larger at the top (outside) than they are at the bottom (the side facing the CCD chip), so they act as light magnifiers and work to increase the sensitivity of the CCD.
Figure 3.2 A CCD image sensor contains three main parts (top to bottom): the microlens, the Color Filter Array, and the photosites. (Diagram courtesy of Fujifilm USA)
Most digital SLRs currently on the market employ sensors that are larger than 1/2'' but still significantly smaller than a 35mm negative. As a result, the effective focal length of SLR lenses is multiplied, usually by a factor of 1.5 or so, when used on a digital SLR. The next generation of digital SLRs will likely use full-frame 24 x 36mm sensors, providing better light sensitivity and eliminating the multiplication factor.
Because many CCD image sensors were originally designed for use in video cameras, they have the same 4:3 horizontal-to-vertical size ratio (called the aspect ratio) as a television screen. Unfortunately, 35mm film has a 3:2 aspect ratio, which is proportionally much wider than a TV screen. Some digital cameras allow you to shoot in either 4:3 or 3:2 mode. The 3:2 mode is very convenient for producing prints on 4 x 6'' paper, because the entire image fits perfectly.
5 x 7'' and 8 x 10'' are also very popular print sizes, even though neither of them match the 3:2 aspect ratio of a 35mm negative. And they don't match a 4:3 sensor, either-although a 5 x 7'' print is a very close match. Keep this in mind when you're shooting an important photo-a group picture at a wedding or other event, for example-that may wind up as an 8 x 10'' print.