Digital Versus Analog Imaging

Abstract

Abstract In nature, the only continuous image is the “true” scene made up of the point‐to‐point reflectance or emission of light from the various objects in the scene or the direct capture of light from a source such as the sun or light bulb or the translucent transmission of light. There is no limit to the fineness of the detail or the amount of light present. Before the digital age, film was considered by many to be a continuous media (to the unaided eye) even though all photograph images were either made up of very small clusters of developed silver or dye clouds that ranged from one micron to several microns in size. For color films, the dye clouds were normally formed in multiple layers (as many as nine imaging layers) and the overlap of the dye clouds gave rise to a very continuous image unless observed under the microscope. Again, before the digital age, film images were scanned by high quality graphic art scanners that created separations (cyan, magenta, yellow, and black) on black‐and‐white film that would then be used to make halftone separations for printing. The resulting halftones looked continuous if the combination of viewing distance and dot pitch was such that the human eye integrated the dots into an apparent continuous image. If the eye failed to integrate the dots, then is was seen as a halftone dot pattern with some image content. When digital cameras and scanners became available at a moderate cost, the notion of a continuous, captured image gave way to the discrete, sampled image. The sampling transformed a region of the continuous image, the pixel aperture, into a well‐defined pixel of some spatial dimension and some well‐defined value (the average over the pixel aperture). The sampling of the image introduced aliasing artifacts (low frequency banding) when the sampling distance was less than the spatial frequency content of the image. The value of the data defined by each pixel is then quantized into some well‐defined number of bits and a calibration curve. Hence a transmittance or reflectance signal between zero and one would be quantized to 8 bits or more resulting in levels between 0 and 255 or higher. In the same manner, electrical signals captured by a digital camera would be quantized, compressed, and stored in the camera storage media. The impact of this discrete sampling and quantization is covered in this chapter.