If you are going to put the image on a web page you want to think about resolution in terms of a computer display. There are several popular resolutions for a computer display and they result in a screen image that is either 800 wide by 600 tall, 1024 by 768, or 1280 by 1024. This means that in a web browser such as Internet Explorer, an image around 750 dots wide will fill the width of the lowest resolution screen (leaving some room for window borders). Aiming at the lowest resolution screen means that the users of these screens will not have to scroll your image around to see it, but also, of course, that it will not fill up a higher resolution screen.

So if you want your 5" wide photograph to fill the computer screen, you could scan it at 150 dots per inch, and end up with an image that is 750 dots wide or nearly enough to fill the lowest resolution computer screen. You don't want to scan at 200 dpi and get a 1000 dot wide image to fit in a 750 dot wide space because it will display much slower due to the larger file size, and the quality will be much less because the web browser will reduce the size and not do it very well.

Many people put images on the web that contain much more data than can be displayed, so the browser reduces them to fit in the desired space, which also reduces their quality, and the file size is so large it takes so long to download and display that many viewers will just give up and move on somewhere else. These images would function better overall if they were reduced in size to begin with.
 

If you plan to print the image you will want to think about scanning in terms of what the printer is capable of, and since most current printers can print at 600 and more dots per inch you want to scan the image at a higher resolution than for a web page, but not necessarily at the full resolution that the printer is capable of, because the file size of the scanned image may become too large to easily deal with. Again, let's assume you are scanning a 3" x 5" photo and want to print it full size.

Due to the relatively high dots per inch capability of a printer, you can usually scan at a lower resolution than you will print at and not notice much reduction in print quality. Scanning the 3"x5" photo at 300 dpi and printing it 6"x10" means that the image will be blown up in size during the printing process, which will cause each dot in the image to become two dots wide by two dots tall. Scanning at 200 dpi and printing 6"x10" means that each dot in the image will become three dots wide by three dots tall. It is this doubling and tripling of dots that reduces quality of the print-- the more an image is blown up to print, the lower the quality. One other factor that affects image quality when changing its size is the number of colors in the image. An image stored in millions of colors may look okay blown up in size four times, but one stored with only 256 colors will become very grainy and low quality as it is blown up in size. The difference is the number of colors; more colors means more shades of color and edges that are less defined and harder to see.

Since we are thinking in terms of balancing file size versus print quality, an image that takes up most of a printed page should be scanned at a much lower dpi than an image that only takes up a small portion of the page, simply to reduce the file size to something that is manageable. A large image that is scanned at a high resolution creates a file size that is so large it will either print too slowly or may fail to print at all.

Suppose, however, that you only want to scan and print a one inch area of a photograph, say a person's head. Scan the small area at a higher resolution (600 dpi) and you will get an image that is 600 dots wide and could probably be printed four inches wide with acceptable quality.

We talk about dpi in relation to scanning the image and in relation to printing, and these are two separate functions. Just because an image was scanned at 600 dpi doesn't mean it has to be  printed at 600 dpi, it only means that 600 bits of information were stored for each inch of image.
 

There are many formats for saving an image in a file, and they are denoted by their file name extensions: TIF, PCX, GIF, JPG, to name a few. The two formats used for Web pages are GIF and JPG.

GIF files can only store 256 colors. Some images, such as photographs, cannot be displayed well in 256 colors and you will see them as blotchy or dotted, but 256 color storage produces a relatively small file.

JPG files, like TIF, PCX and several other formats, can store millions of colors, which usually produces a very large file because each dot is represented by a number between 0 and, say, 64 million (depending on the color setting). It takes more file space to store the color information in the number 58,333,245 than it did to store the number 256, so the file gets much larger than a GIF. The JPG file, however, reduces file size by throwing away a certain amount of color information; the amount discarded is controlled by the quality setting on the SPG file save routine. The lower the quality setting, the more color information is thrown away, and the smaller the file size. If, for example, 10 dots in a row are a very slightly different shade of Red, the JPG compression may save them all as the same color to save space. JPG compression is called "lossy" because it loses information to save space. A sample photograph created a file of 950K as a compressed TIF, and 32K as a medium-high quality JPEG. Different JPEG saving routines use different quality/size settings. When saving from Photoshop to JPG a setting of 3 out of 10 ( lower than middle quality) can be used for web work where we sacrifice some quality for smaller file sizes. Saving from the scanner software requires a setting of 3/4 of the way up the slider, which is higher than middle quality.

JPG files saved in medium to lower quality are the best choice for complex colored images on Web pages because they have a lot of color information and are fairly small in size. GIF files are fine for buttons, graphs and other images that do not require complex color.

If you are working on a graphic image in an editor keep this in mind; each time you open and save a file in

JPG format, more information is lost and the image quality goes down. If you are editing a graphic it is best to work on it in a "non-lossy" file type such as TIF (with compression turned on to save space), or if you are using Photoshop, the native PSD format is fine. When you are finished with your editing, save out the copy you will use on the Web in JPG format at a fairly low quality setting-you still have the high quality original to return to if necessary.

NOTE ON PHOTOSHOP IMAGE SIZE SETTINGS