The short version
When you scan an image it is important to consider what you intend to do with the resultant file. Do you want to:

* Print it as a high quality enlargement?
* Use it in a Word file?
* View it on a computer screen?
* Send it as an email attachment?
* Include it as part of an email?

Unless you are very careful, a file intended for producing high quality prints is not suitable for sending via email, and may cause difficulties when viewed on a screen. Similarly a file that works perfectly via email will look awful when printed.

Working from a typical postcard sized photo, an "average" scan would be at 200dpi, with normal compression to give a JPG file. Such a scan will produce a file with a size from 150 to 300K; when viewed in Internet Explorer it will appear two and a half times its original size (bigger than your screen); it will give a fair quality print; and some ghosting will be visible.

Print quality can be improved by increasing the resolution (higher dpi, less compression), but this will give a larger file.

* To email family snaps to friends overseas then use a JPG file at 80 or 100dpi with normal compression.
* For classy photos to be printed in advertising brochures or as part of a professional report then use a JPG file at 300 or 400dpi with minimal compression.

An Image in a File
A photo can be stored in a computer file. This can be done by using a Digital Camera or by using a Scanner.
An image in a file is made up as lots of Pixels (Picture Elements). Imagine that you place a piece of fine screen mesh over a photo so that the holes in the mesh form a grid of tiny squares. Each of these tiny squares could be treated as one Pixel and the average colour characteristics of each little Pixel can be expressed as a number.The Resolution, or quality, of the photo stored in the file will depend on:

* how small the Pixels are. Historically this is measured in "dpi" (Dots Per Inch) the more dpi the bigger the file, but the better the resolution.
* the kind of number the colour characteristics are stored in. This depends on the Mode and Bit Depth used and type of file created (e.g.: JPG, TIF, BMP).
* any fancy "Compression" tricks that may be used to try and make the file smaller. The less compression the bigger the file, but the better the resolution.

Typically, the number to store each Pixel uses four bytes (characters) and since there is one number for every pixel, an image file can be very large. Consider that a postcard photo is 150x100mm or 6x4 inches. At 200dpi we will have 6 x 200 x 4 x 200 pixels (960,000 pixels). At 4 bytes each this gives a file of 3.84 million bytes. The same photo at 100dpi would be a quarter of the size. Various tricks and compression techniques can be used to make image files smaller, but most of these result in some loss of image quality.

Generally speaking, the bigger the file, the better the printed result. Unfortunately bigger files can cause storage problems if you try to save them on a floppy disk, or send them in emails.
The best resolution characteristics to use will depend on what you intend using the file for.
Screens and Printers
An image on a computer screen, or a TV screen for that matter, is made up of lots of little dots. One set of dots is used for each Pixel (short for Picture Element).

Colour screens have a blackish background and use RGB (Red, Green, Blue) technology; there are three dots that make up each Pixel - one each for red, green and blue. These dots are very close together, sitting on a black background. By varying the intensity of each red, green or blue dot the screen can fool your eye into seeing any colour in the spectrum.

On a typical screen these pixels are arranged in a grid with 75 pixels per inch (or about 30 per centimetre) across and down. Older black and white screens only had one dot per pixel, hence the term Dots-Per-Inch or "dpi". So we say that a typical computer screen displays pictures at 75dpi.

These figures depend on Control Panel settings and some screens can do much better than this.
An image printed on a piece of paper is also made up of lots of little dots, but since the background is white, colour printers use CMYK technology where the dots are Cyan, Magenta, Yellow and Black.

The dots on printers can be various shapes and may overlap each other. The principle is the same; the C, M, Y or K dots at any point are referred to as a pixel and varying the intensity of these four tiny dots of colour at each pixel can fool your eye into seeing any colour in the spectrum.

If an image is printed at 75dpi it will appear to be very grainy or "dotty" and you can easily see jagged edges where there should be smooth curves. People put up with this on screens, but they expect better on printed documents. Most people can just detect jagged edges at 200dpi, but not at 250dpi.

An image may look good on a screen at 75dpi, but to print it should be 200dpi or better.
So what is Resolution?

The "Resolution" of an image is an expression of it's clarity. This can be effected by the number of pixels (dpi); the depth of colour (the amount of space allowed for the number that represents each pixel); and the amount of compression allowed. Generally, bigger files give better resolution; but sometimes the size of the bigger files causes more problems than it is worth.

Digital camera files

Digital cameras express their resolution as two numbers representing the number of pixels across and the number down a photo. This may be followed by a code (E, N, or F) to represent the amount of compression allowed. e.g.: 1024x640N is a photo 1024 pixels wide and 640 pixels down with Normal compression. On a screen at 75dpi the photo would measure about 13.6x8.5 inches (35x22cm). But printed at 300dpi the photo will only measure 3.4x2.1 inches, a quarter of the screen size.

A question of size

If you have a simple image file and open it in Internet Explorer or an email program, it will appear on the screen so that each pixel in the file occupies one pixel on the screen. A 2048x1280 digital photo will appear so large you will only see a quarter of it on the screen.
But if you insert the same file as a picture in Microsoft Word it will only appear as big as the current margin settings allow, and in any case you can simply make it any size you like.
In this case the software processes the grid of pixels from the file and mathematically modifies them to fit into the grid of pixels on the screen. Individual pixels on the screen are the result of averaging the corresponding pixels from the file.



Glossary

Pixel: An abbreviation of Picture Element. A computer considers a photo as being made up of a "grid" of lots of tiny Pixels. It uses a number to record the average colour characteristics for each and every Pixel.

Dot: An historical term that refers to the spot occupied by one Pixel on a screen or printer.

dpi: Dots Per Inch. The number of Pixels per inch both across and down. At 300dpi a 6x4 inch photo will contain 2,160,000 Pixels. Bit Depth: Determines the size of the number used for recording colour details at each Pixel. 1 Bit allows Black or White only, 8 Bit allows 256 different colours, 32 Bit is virtually unlimited.

RGB: Computer and TV screens use RGB technology. By displaying varying intensities of Red, Green and Blue light on a black background at each Pixel they can simulate any desired colour.

CMYK: Printing devices use CMYK technology. By printing varying intensities of Cyan, Magenta, Yellow or blacK on a white background at each Pixel they can simulate any desired colour.

Compression: Graphics files can be very big. Compression techniques may be used to "squash" the data in these files so they take less space. Most compression techniques will degrade the quality of a photo.

JPG: (or JPEG) A common file format used for graphic files. For general purpose use, works with most programs. The best choice if you have no reason to use some other format such as TIF, GIF, BMP etc.