As a Luddite imagery guy of the paleo-LANDSAT era, a full description or specification of an image usually included the number of pixels (a 1k x 1k image contained 1 million pixels). It also mentioned the depth or graylevel resolution of each pixel ( BIT = 1 graylevel per pixel--black or white, BYTE = 8 bits = 256 graylevels per pixel, SHORT INTEGER = 16 bits = 32,768 gray levels, and so on...). Also included in the spec was the number of bands; 1 band was black and white or panchromatic image, 3 bands could be assigned to the Red, Green, Blue color guns of a true color display device, and usually additional bands assigned to the IR or UV could also be registered to the image.

So, for example, a typical image suitable for a fairly realistic magazine color illustration of a portrait, still life or landscape might have 3 bands (RGB) of 512 x 512 byte pixels: a total of 3 x 512 x 512 x 8 = 6,291,456 bits.

I'm not being nit-picky here, I'm just curious to see how much information can be extracted from one image, its intensity, color, and spatial resolution--how faithful its color, how much magnification it can stand, and the full dynamic range of its light levels. The example I gave above would be fine for a color snapshot of a pretty girl or a sailboat, but would stand up to only limited enhancement, or be of limited use for scientific analysis.

It appears the nomenclature conventions have changed since I worked in the field.