A few scanning tips


Never used a scanner before?

For those who have not worked with a scanner before, here is a very basic introduction to get started, a brief overview of how it works, how you would actually "use" a scanner.

The flatbed scanner is very much like a copy machine, to the extent that it has a glass plate under a lid, and a moving light that scans across under it. Except that scanners can do great color too, and have more controls possible. And like a copy machine, a flatbed scanner allows you to scan photos, paper documents, books, magazines, large maps, or even 3-dimensional objects (those that don't have a lot of depth, coins for example), etc. It scans very much like a copy machine. But instead of creating another piece of paper like a copy machine does, we instead create an image in memory, which we can do with as we please. We have acquired a digital image, and we can show it on the screen, or write a file and email it, or print it, anything we wish.

We must use software to operate the scanner. Some scanner software can operate by itself, and some cannot. Most image editor programs can operate the scanners software (like for example Adobe Photoshop or Elements, Paint Shop Pro or PhotoImpact (both are Corel now), Picasa, Irfanview, Faststone, etc, etc. (NOTE: The current Photoshop CS6 version requires the 32 bit version to operate the 32 bit TWAIN software, the 64 bit version of Photoshop cannot).

These editors have a menu at File - Acquire or File - Import or File - Scanner that you use to scan an image. Sometimes there is a toolbar Scanner icon too. This menu starts the scanner's software, called a TWAIN driver. A fancy word, it is really just another program, but TWAIN is a software standard that all the scanner manufacturers agree to use, intended so that all image programs can operate all scanners. It wasn't always that way, not so long ago you could use only the one image program that came with the scanner. Today you have your choice of most photo editors.

Chatty... TWAIN is often said to be an acronym for Technology Without An Interesting Name (humor), but the FAQ at the www.twain.org organization says it is not an acronym for anything. They say it instead comes from Rudyard Kipling's line "East is East, and West is West, and never the twain shall meet" (The Ballad of East and West, 1889).

The TWAIN driver itself is simply the software provided by the scanner manufacturer to operate the scanner (the user interface), and it is also the interface between the graphics program and the scanner hardware. Twain refers to that separation, now photo editors don't have to know how to operate scanners. The Twain driver will always have another name too, Microtek calls theirs ScanWizard, Umax has two, VistaScan and MagicScan, and HP had a version called DeskScan, and the current is called PrecisionScan. The TWAIN driver comes with the scanner, and knows how to operate this one brand of scanner hardware. Each scanner manufacturer provides their own TWAIN driver for their hardware. You can use any image program, but you typically must use the provided TWAIN driver.

From our human viewpoint, the TWAIN driver ordinarily has the controls that we use to specify the scan; we set the mode (Color or B&W or Line art), we set 100 dpi or 200 dpi resolution, we set the area on the flatbed glass that we wish to scan, and there are controls to help correct the tonal quality and color balance of the scanned image. After the scan completes, the TWAIN driver transfers the image into the image program memory. That is, the new image just appears back in the image program. Then you can tweak the image more, or print it, or save it to a disk file.

The scanner would be used like this:

Start the photo editor program.

Select the FILE - ACQUIRE menu, or often there is a toolbar button for it. Some software calls it FILE - IMPORT instead, same thing, it is just to keep you on your toes.

Here is a look at a few samples of menus of various popular image programs that shows the Acquire/Import menu used, and how to find it.

Usually there is also a submenu called SELECT SOURCE where you select your TWAIN driver (called TWAIN32 now). You might have two scanners and you will see both of them there. You must make the scanner selection the first time, but only the first time, until it changes. Then you select Acquire again, and Select Image, and the scanner's TWAIN driver will start and appear.

Microtek ScanWizard looks like this, and others are similar in that there are Preview and Scan buttons, and a Preview view of the glass bed.

The first step is the Preview button. When you press the Preview button (or Overview in this case), you get a quick overview scan of entire scanner bed, or as much of its length as you have previously selected (at the Preferences menu). The Preview is a quick low-resolution scan, and that image of the scanner bed appears in the Preview Window. The units along the top and left edge are inches. This is a 3.5 x 5 inch photo shown on the 14 inch bed of this flatbed. Often there are magnification tools, but the final scan is the important view.

To describe the position of the subject on the glass bed to the scanner, you use the mouse to mark a rectangle on the image of the bed, describing the area you want to scan (that's the dashed line shown). You can mark the entire photo area, or can crop the image to be scanned, by marking only a part of it, or perhaps just a 1-inch square area from it. The Scan button will then scan only the area that you have marked with the mouse.

TYPE is shown selecting the Color Mode, see below for Type. You specify the Scanning resolution here too. The 300 dpi shown creates a large image, but you specify whatever you need for your purpose, normally 75 to 300 dpi is common, but purposes do vary (2400 dpi is used to enlarge film — forget all about using 2400 dpi to scan photo prints). Scanning resolution is specified in dpi or ppi (both terms same exact same meaning, Dots Per Inch means Pixels Per Inch).

A Summary

Trying to be brief here, for those unwilling to read anything. This part is very important Basics.

There are generally two different goals for scanning an image, to show it on the video screen, or to print it on paper.

The purpose of high resolution is for enlargement, so an important way to see it is this way:

Print Resolution:

Scanning and printing at the SAME resolution will print an original size copy. 300 dpi is a good goal for color photos, or 600 dpi for line art or B&W text documents.

The ratio of (scanning resolution / printing resolution) is the enlargement factor.

Scan at 600 dpi, print at 300 dpi, for 600/300 = 2X size (double size or 200% size)

Scan at 300 dpi, print at 300 dpi, for 300/300 = 1X size (original size or 100% size)

Scan at 150 dpi, print at 300 dpi, for 150/300 = 1/2X size (half size or 50% size)

Don't go crazy with the resolution. If you scan a 8.5x11 inch page at say 4800 dpi, you will get a 6 GB image, which is absolutely the last thing you could ever want. Instead you may want to PRINT a 8.5x11 inch page at 300 dpi (about 8 MB image) ... a result to fit your goal, to match your intended use for the image.

But film is small, and does need more enlargement (more scan resolution). Where photo prints have already been enlarged, you probably want an same-size copy.

One example is to scan 35 mm film at 2700 dpi, and print at 300 dpi, for 2700/300 = 9X size enlargement. 9X is about 8x12 inches (about A4 size) from full frame 35 mm (about 0.9 x 1.4 inches). The ratio of (scanning resolution / printing resolution) is the enlargement factor.

This enlargement concept is true for scanning anything, photo prints, documents, film, etc.

Video Screen Resolution:

Video screens are Very different than paper. Paper is dimensioned in inches (or mm). Screens are dimensioned in pixels (no inches are involved on the video screen). And images are dimensioned in pixels. The role of scan resolution for the video screen is to create the appropriate image size.

For one example, scan 6x4 inches at 100 dpi, which will create (6 inches x 100 dpi) x (4 inches x 100 dpi) = 600x400 pixels. These 600 pixels will fill 50% of the width of a 1200 pixel screen.

Plug in the appropriate numbers to get the size you want from the size photo that you are scanning.

Your video screen size might be set to show say 1024x768 pixels (this varies, we can set different size screens), and the concept is that this 600x400 pixel image will fill 600x400 pixels of that 1024x768 pixel screen (this example image width and height is slightly more than half of this full screen size).

That formula is true for printing goals too, meaning that if you will print 8x10 inches at 300 dpi, then in preparation, you need to create in the ballpark of (8 inches x 300 dpi) x (10 inches x 300 dpi) = 2400x3000 pixels

Here is an enlarged list of basics and differences. Here is a Scanning and Printing Calculator.

This resolution number determines the number of samples (pixels) taken from the photo, and therefore the size of the image you will get, like 800x600 or 300x200 pixels. If you scan 6 inches at 100 dpi, you create 6x100 = 600 pixels. This is covered in the Video Resolution Basics section. You can change the units from Inch to Pixels or Cm to see the final image size in pixels or cm.

The Settings screen above shows that we selected 4.92 x 3.42 inches on the photo on the flatbed glass (the red dashed line box shown). At 300 dpi, this will create an image size of 1476x1026 pixels (this size can be shown if units are switched from inches to pixels above). This size image will consume 4.4 MB of memory (also shown).

If we wanted to scan say 4.5 x 3.0 inches instead, we could simply overtype the size numbers here to get it, or we could remark with the mouse. The resolution multiplied by the size of the marked area determines the image size in pixels (see the Video Resolution Basics section). We can change the displayed units from inches to pixels, to see and predict the image size in pixels on video screens. This is all covered in detail later in the Basics 101 section.

There are many controls you can use to improve the image, but for now, I'm trying not to confuse things with details. The additional controls provide a way to get even better scans. In ScanWizard, the tiny button at the right end of the lower toolbar reveals a large RESET button that sets all the defaults, to achieve a known starting point for the scan.

Then you set the TYPE of scan (above). Scanner modes offer the following choices to get image types illustrated by these 50 dpi scans of a small photograph.

All scanners should have these three standard modes, Color, Grayscale, and Line art (but some scanners might rename the modes).

Color - "Millions of Colors" above. Scans three color samples per pixel, one for each Red, Green, Blue (RGB) shade, the combination of which is the one resulting color from the 16.7 million color choices possible with 24 bit color. This 16.7 is computed as 256 shades of each Red, Green, and Blue, which will likely actually total no more than several tens or hundreds of thousands of colors actually present, out of the theoretically possible 16.7 million color combinations.
Grayscale, 256 shades of Gray - Like a Black and White Photograph, for scanning photos, pencil drawings, etc. 8-bit Grayscale uses only one byte per image pixel, but 24-bit color uses 3 bytes per pixel.

Some scanners might rename this mode to call it "B&W Photo" mode. Only shades of gray, no color.

Line art - only 2 colors, either White or Black. This is the same as fax, or clipart, i.e., no gray. The name comes from ink drawings. It is a very poor choice for photographs on the same page, when it is much too contrasty, and only minimal photo detail can show.

Some scanners may rename Line Art mode to call it "Black & White" mode instead. Exact same thing. Only two colors, black or white.

Line art - However, Line art mode is ideal for pages of printed text, which might look like this. This is 12 point text scanned at 300 dpi from a magazine. It would appear normal size when printed on the printer.
There are ways to improve such line art images with software, called dithering. This image was scanned as Grayscale, and was then dithered to 2 color B&W, and could be faxed. Other than fax, you'd need a good reason not to use one of the better modes for photos. This one could be better if it were larger.
Halftones - this scanner mode is for images to appear in magazines and newspaper pages. However, this mode is normally NOT used at home, and halftones are never used for video screen images. Our printer drivers instead provide default halftones that are normally used instead. This is a 4x4 halftone screen, and it needs a much larger image, matching the Line art printer resolution at 600 or 720 dpi to look right. Instead, it's only 50 dpi here for the video screen size.

To continue - You press the Scan button and the scanner will scan. It takes 15 to 60 seconds perhaps, depending, and then the image will automatically appear back in the original image program, the same as if you had opened an image file. You have now "acquired" an image.

You can modify this image in the image programs if desired, lighten it, darken it, crop it, resize it, change its color balance, touch up small areas, edit it drastically, apply various filters like sharpening filters or descreen filters, etc, etc. Or with some TWAIN drivers, you can do all these things as you scan. Often you need to do nothing more if you got it right in the driver. And, you can always discard this image, make a small change, and see the next try in a minute or so. It's not at all unusual to scan an image a couple of times before getting it exactly right. It depends on the purpose and how critical it might be, and if you keep getting better ideas. 😊 But the beauty of a flatbed scanner is that the photo lies motionless on the scanner bed all this time, it doesn't move, and you don't have to keep marking the crop area for every scan. Nothing changes except the settings that you change, so you simply tweak the control, and hit the Scan button again. Nothing to it.

When the image is as you want it, then you can either print it or write it to a disk file, or both. You would use the FILE - SAVE AS menu to specify a file format like TIF or JPG, and a file name. Then you have an image that you can put on a web site or email to someone (see here about how to email an image).

(Continued at the NEXT button below)

Copyright © 1997-2010 by Wayne Fulton - All rights are reserved.


Main Next