At the heart of today's digital imaging devices are CCDs (charge-coupled devices). A type of semiconductor that's sensitive to light, a CCD consists of a two-dimensional array of individual elements, each of which is, in essence, a capacitor — a device that stores an electrical charge — thus explaining the first and last letters of the acronym.
A CCD's charge is created when photons strike the semiconducting material and dislodge electrons. As more photons fall on the device, more electrons are liberated, thus creating a charge that's proportional to the light's intensity. With a 2D array, you can capture an image.
Put another way, each CCD represents a single-image pixel. Today's best digital still cameras have sensors with up to six million pixels.
The challenge lies in reading these charges out of the array so they can be digitised. To do this, each individual CCD detector, or pixel, consists of three transparent polysilicon gates over a buried channel of doped photosensitive silicon that generates the charge. The channel is flanked by a pair of channel stop regions that confine the charge.
To read and digitise a particular CCD's charge, the voltages of the three gates are cycled in a sequence that causes the charge to migrate down the channel to the next gate, then to the next pixel, and ultimately down the row until it reaches the end column, where it's read out into a serial register and ultimately sent to an ADC (analogue-to-digital converter).
Think of this process as something like a bucket brigade, where water in a bucket at the beginning of a line is transferred to the end of the line after being passed from bucket to bucket. This charge transfer occurs with an efficiency greater than 99.9 percent per pixel.
The sequence of moving the charge from one gate to the next is called coupling (which explains the second C in CCD).
Coaxing out colour
But after that's all said and done, the CCD imaging array is only sensitive to light intensity, not colour. One way to capture a colour image is to use three CCD arrays, each covered by a filter (usually produced by painting the CCD's surface with dye) that passes one of the three primary colours — red, green or blue. Onboard camera electronics merge these primary components into a colour pixel. Because it requires three CCD arrays, this system is found only in high-end cameras and camcorders.
A low-cost method applies a special colour grid, known as a Bayer pattern, over the imaging array. This pattern of alternating red-green and green-blue filters enables a single CCD array to capture a colour image. Half the filters in this layout are green because the human eye is most sensitive to that colour.
The CCD, invented in 1969 at Bell Labs (now part of Lucent) by George Smith and Willard Boyle, was originally intended to store computer data. But that function was taken over by faster technologies. By 1975, CCDs were being used in TV cameras and flatbed scanners. In the 1980s, CCDs appeared in the first digital cameras. CCDs are widely used today, but they do have some drawbacks.
Why film is still better
Although the coupling process is quite efficient, moving the charges along a row of many hundreds or thousands of pixels adds up to a noticeable loss of charge, which leads to fading. Also, if too many photons strike a CCD element, it gets 'filled up', and some of the charge leaks to adjacent pixels, creating blooming. If light strikes the sensor while a transfer is taking place, it can cause some data loss and leave streaks behind bright areas of the image.
Finally, CCDs require a different manufacturing process from other computer chips (such as CPUs and memory), so specialised CCD fabrication plants are necessary.