There are some key issues about scanners and printers that should be understood by anyone who wants to master high-quality digital imaging and printing. These issues or factors are: color depth, dynamic range, and resolution.
color depth
Color depth is just another way of saying bit depth for scanners, and the same principles apply. A normal color original (print or film) will require a minimum 24-bit scan (8 bits of information per RGB channel; all scanners scan in RGB) to reproduce it with adequate fidelity. This is the old “millions” scanner setting. (Most modern scanners are 12-bit scanners. Even some that say they are 16-bit actually only use 12 bits of data carried in a 16-bit format that is more efficient for computers.) Aren’t these millions and billions? simply excessive colors? Perhaps, when you consider that until recently, with the introduction of Adobe Photoshop CS, hardly anyone could edit or work with these high-bit files, and as a result, they were always downsampled or converted to 24-bit images by the scanner. However, the advantages of adding more information up front (more raw material or “margin” to work with) at the scanning stage have become more obvious to people, especially those working with film compared to those scanning prints. where the range of density is much lower. .
Note that many high-bit scanners still output files in 24-bit (some produce true 48-bit output). Why only 24 bit output? Because humans can’t really see 48-bit, and also because computer monitors operate in 24-bit mode (although most digital printers can accept 48-bit data, which is usually converted to 24-bit on the fly). However, you still want to scan in high-bit mode because a 36-bit scanner has more steps in the density range between deep shadows and light highlights than a 24-bit scanner. (Actually, there aren’t any 24-bit scanners anymore; they’re 36-, 42-, or 48-bit.) A 24-bit scan means 8 bits per RGB channel, which equates to 256 possible brightness levels per channel (0-255, where 0 is pure black and 255 is pure white). A 36-bit color scan means 12 bits per channel or 4096 possible values. That’s a lot more tonal possibilities. And this is especially important in the dark shadows of positive or inverted film (which with negatives become reflections) where you want as many steps as possible to pick out unique details. Shadow detail is often what makes or breaks an image. The downside, and there is always a downside, to scanning and then editing bit-heavy image files is that increasing the bit depth increases the file size arithmetically. However, as scanning expert David Coons says, “double the file size is a small price to pay for the 256 times increase in luminance accuracy achieved by switching to 16 bits per channel.”