Digital revolution looms from blur
The digital revolution in photography nears a significant threshold - cameras which will push image resolution to the point where acceptable 8x10 inch prints can be made.
To gain a better appreciation of what is so significant about where we are in digital scanning technology, an explanation of how digital cameras make images, and how the quality of images is measured, is in order.
In a conventional camera, film captures and stores an image. In a digital camera, these functions are separated - a part called the CCD captures the image, and a memory chip stores it.
Film scanners, flat-bed scanners and video cameras all use CCDs. But there is substantial difference between the image quality from, say, scanning a 35mm transparency using a $10,000 Nikon film scanner and an image shot with a similarly priced digital camera.
How these devices utilise CCDs varies considerably. Of the three - scanners, video cameras and digital stills - the CCD in the digital still works the hardest.
CCDs are far from perfect image-capturing devices. There are colours they can capture which film cannot - greens in particular tend to be stronger.
But there are many more colours CCDs cannot capture. Blues tend to be weak, something that causes problems in the wrong lighting situation.
CCDs also have trouble capturing the darker parts of a scene.
Computers work in something called RGB colour. The CCD captures the reds, greens, and blues of an image separately and then recombines them on a monitor or TV screen.
When a scanner captures an image, it passes over the original using three coloured lights to illuminate the reds, blues and greens. Early scanners made three passes, and some still do, but most now make a single pass with one light following slightly behind the other.
As the light illuminating the original is controlled, the user can best exploit the strengths of the scanner's CCD and get the best possible image from the machine, something that cannot be done with a digital camera which is limited by the lighting conditions under which you are shooting.
It is the three passes that make the difference in colour quality. Since the original is scanned three times, the entire resolution of the CCD is utilised.
That is not the case with a digital still. Since each colour must be captured in a single exposure, the camera's CCD must be divided into three parts, one capturing the reds, one the greens and one the blues. This is done by making the first pixel on the chip see only reds, the second blues, and so on.
Of course, the actual resolution of each channel will be what the manufacturer claims, 2.2 million pixels for example. But within each of the RGB colour channels, the actual resolution is one-third of what it would be if the image were made by a scanner, with the remaining information fudged in. Thus, a scanner using a similar quality CCD can produce much more realistic colour.
High-end digital cameras get around this by using three exposures to capture an image - meaning the cameras are good only for still-life images as they must shoot each subject once for reds, once for greens and once for blues.
The quality difference is tremendous. High-end, three-shot digitals are considered by many as superior to traditional film-based counterparts, not just in turn-around times, but also in image quality.
There is a third type of image capture device - the video camera. There basically are two types of video cameras - the cheap ones and the expensive ones.
Even cheap videos have a bigger advantage over a digital still camera with the same CCD.
Since you are looking at 30 still images per second on one video segment, many of the flaws in each frame are covered as one frame blurs into the next. With the high-end models you have three CCDs, one for each colour, so the colour quality is much better.
This design is coming to low-end digital camera, and soon we will begin to see much better image reproduction. Then comes the revolution.