How much detail do I need? How big a monitor on my computer? How many pixels in my digital camera? (A pixel is a picture element - the tiny dots that make up a digital picture - but each dot is a single solid color.)
Can you recognize the picture in an icon on your computer desktop? Probably. But would you want to make a full page print of that picture to post on the wall? Unlikely. (An icon typically has about 1,000 pixels, an ID photo has about 5-10,000 pixels.) The amount of detail that is required in a picture depends on what you are going to do with it - or in simple terms, how big it will be when you use it.
One of my favorite digital cameras supports the following resolutions:
VGA is basically broadcast television quality, and twice the resolution (four times as much detail) as most camcorders. A snapshot size photo (3 1/2 by 5 inches) is "pretty good" using that quality. So we could say that 307,200 pixels does "pretty good" giving us 17.5 square inches of picture - about 17,500 pixels per square inch. Or a resolution of roughly 128 to 137 pixels per inch on the print.
Most home printers use 8.5 x 11 inch paper. So this has become the "normal" size home picture, rather than the 3.5 x 5 inches from the drugstore photo lab. And the digital camera folks said that you needed a 2 megapixel camera to get a good quality print at 8.5 x 11 inches. Were they just trying to sell us a better camera? Do the math. 935 square inches of picture, with (using the options above) 1,920,000 pixels, is 20,535 pixels per square inch - maybe enough better than the "pretty good" numbers for snapshots that we could actually call it "good." Or a resolution of roughly 141 to 145 pixels per inch.
The camera companies also suggested that at least 3 megapixels were required for professionals, who often have to produce 11 x 14 inch pictures. Let's do the math again. 154 square inches of picture. 3,145,728 pixels (from the list above). 20,427 pixels per square inch. 140 to 146 pixels per inch of paper. Totally consistent with the previous example.
But there is a catch. What if I want a snapshot-size enlargement of just one face from the shot of the family reunion? Or just one building from the vacation scenery. If I only want 1/3 of the picture (in each direction), I only have 1/9 of the pixels to work with. If I want the same quality I would have had if this was originally the complete picture, then I need 9 times as much data (since 8/9 will be discarded). If 300,000 pixels are required for a good snapshot size picture after cropping, then 2.7 million pixels are required in the original picture before cropping.
As digital electronics has gotten better and cheaper, digital cameras have expanded as well. If 3 megapixels is good, isn't 5 megapixels better? With a good lens and other components, perhaps it will help. But the storage required is very high. Nobody will want to see those pictures. In practice, you may want to keep the "original" just as we have traditionally kept the film negatives. But after cropping and adjusting the picture, I recommend that the resolution of your working copy be reduced to approximately 300 rows or columns for display in email or web pages. Your viewers will thank you.
One expert estimated that a 35 mm negative shot on traditional 200-speed film can produce the equivalent of 18 million pixels of resolution - far more than an ordinary digital camera, which is up to 3-5 million pixels. A skilled photographer with plenty of time might get a great picture with a camera that provided 1 million pixels, but few people are that good.
If you have a good camera (sharp pictures) and are a good photographer (your pictures don't require major cropping), use the XGA resolution above - or whatever you have available that gives between half and one million pixels. You will easily be able to make snapshot prints, and can do some larger prints. The pictures will store and transfer quickly. (They will normally compress to under 200,000 bytes)
If your camera isn't sharp or is sometimes out of focus, or if you need to do extensive cropping to get a good picture, keep more resolution in the original photograph. After you are done editing, you can reduce the resolution to make your photos take less storage. But if you start without the detail, you can never get it back if you want it later.
If you are an active user of document imaging, you should be able to look at a full page at a time - not waste time scrolling around the page. The page should be life size (many forms already use "fine print" as small as you are expected to be able to see, so a smaller version of the page won't be adequate.) And you should have enough detail to be able to recognize the fine print without zooming in.
If you are just an occasional user of document imaging, almost any size monitor will work. Your productivity may be a little lower, because you may have to zoom into a page (view part of the page at a time), but by definition, this is not a lot of time each day.
CRTs and other displays are measured diagonally (early CRTs were round, so the standard measure was the diameter, which corresponds to the diagonal in today's tubes). A 21 inch display has just over 11 inches useful area vertically - factor in the window border, and you still have a life size office document. So for most office document users, a 21 inch monitor is recommended. Those monitors are wider that 8 1/2 inches, leaving room for the other programs that must often be used at the same time as the image. Can a smaller monitor be used with document images - certainly, but it may be necessary to only show a portion of a page at a time, thus reducing the efficiency of the operator.
How much detail is required? In the days of dot matrix printers, 7 dots vertically could barely show the shape of each character, and could not represent any style to the font - those printers were hated and replaced. 9 dots was about the minimum acceptable to recognize the text clearly, and even that was in the ideal world where the shape of the letters could be adjusted for best use the dots/pixels. Add another "pixel" for the space between lines, and it takes a minimum of 10 pixels per line of text. Normal text is 10-12 points (7.2 to 6 lines per inch). So when we were using dot matrix printers normal text required at least 72 pixels per inch. With a scanned images, we don't control the shape of the letters, so need around 100 pixels per inch. 100 pixels per inch on a screen just over 11 inches tall corresponds to 1600 x 1200 resolution - the most often supported and used resolution for 21 inch monitors. Users with monitors smaller than 21 inches should still use about 100 pixels per inch - for example 1280 x 1024 on a 17 inch monitor (about 10 inches high).
What about fine print that may be 5 or 6 points - half the normal size? At 100 pixels per inch, each row of text is just under 7 pixels high. Since the text may be any shape, not optimized for so few pixels, we may be able to recognize this small print "in context" - know the labels on the form say "home address" or "business address." But if we need to read a telephone number, and recognize every character "out of context" from a font that is half the normal size, then we need twice as many pixels per inch. Therefore I recommend scanning and storing office documents at a minimum of 200 pixels per inch. Routine display may be at 100 pixels per inch, but if we need to zoom in and double the size of the image, the detail will be there, not just a larger blurry image.
Early monitors were often 11 to 15 inches (although a few of us older folks even used monitors as small as 5 inches). With monitors that small, each application had to fill the screen - with early Windows and OS/2, people normally maximized each application as soon as it was started.
Many people who get a large monitor follow their traditional procedure and maximize each application. The large screen allows them to sit farther from the monitor, so that the need to enlarge the application is self-fulfilling. The icons on the desktop seem terribly small if you are a long way from the monitor, which leads to further customization of the screen.
Unless someone is vision impaired, the normal reason for a large monitor is to have more information available at a time - e.g. to simultaneously see both the image of a document and the program that collects data from that document. Each program may take as much screen area as it took on the smaller monitor, but now only fills part of the screen.
At first it will seem awkward looking from side to side on the display - but it is like looking side to side to see all the papers on your desk. And the user may need to sit fairly close to the monitor - no closer than you sat to a small monitor, and no closer than you might hold a paper you are reading, but with the large monitor it will seem much closer.
How should you set up a desk with a large monitor that will be used as a primary workspace?
Back to the home page at www.plesums.com
Back to the Document Imaging index at www.plesums.com
Send e-mail comments to Charlie@Plesums.com
©2004 by Charles A. Plesums, Austin, Texas USA. ALL RIGHTS RESERVED. If you would like to make or distribute copies of this document, a nominal royalty payment is required, as specified on www.plesums.com.