DUKE-FAMILY-L Archives

From: "Tony Cox" <>
Subject: DUKE/FOSTER: Scanner How-to
Date: Tue, 1 May 2001 13:58:28 -0500

If you have no interest or need for basic information on scanners, you
may want to delete this note now.

I was recently asked for some how-to information on the use of scanners.
Below are my ideas on scanning techniques.

Scanners are very useful for genealogy studies. I use mine in many ways.
I scan data records to permanently store them on my computer for easy
access, to display them on my online web page, to store the records on
CD-ROMs for my family, to send copies of the records over the internet,
and, with my printer, to make paper copies of them for notebooks and for
other genealogists.

Scanners usually come with graphics software that allow one to make
minor repairs on copies of old pictures. The programs can also be used
to enlarge and to enhance hard to read portions of the records. I used
mine this weekend to help transcribe the smeared ink and bled-through
writing of an old 1809 South Carolina court record. Sometimes character
recognition software will also be supplied with the scanner. With it,
printed text can be 'read' into a word processor file.

Although hand-held scanners have their uses, the best kind of scanner
for genealogy studies is the flat-bed scanner. Flat-bed scanners come
with three types of computer connection. The older scanners usually
required a special interface card and installation involves some
technical knowledge. The more recent ones have either a parallel, also
called a printer, connection or a USB, Universal Serial Bus, connection
or both. It is easy and simple to connect a scanner with a parallel or
USB interface. If your computer has a USB port, most newer ones do, it
is the one I recommend. Otherwise, the parallel one will work fine.

Read the connection instructions that came with your scanner. Sometimes
the instructions ask you to install the software before installing the
hardware. Be sure to remove power from your computer before connecting
the scanner hardware. The USB cable connection is straight forward. The
parallel connection is usually made by disconnecting the printer cable
from the computer and connecting it to the scanner. Use the cable that
comes with the scanner to connect between the scanner and the parallel
port on the computer. The parallel port is where the printer was
connected. In computer parlance, the scanner and printer are said to be
'daisychained' together. Now turn power on. The computer should detect
the new hardware and step you through the installation process.
Otherwise, follow your scanner's instructions. If the scanner
connection is through the printer interface, you may also have to
re-install the printer software.

Scanning an image is a three-step process: preview, adjust, and final.
When you activate the scanner, either through your graphics software, a
menu selection, or a button on the scanner, a preview scan is made. The
scanner program makes a 'best guess' of the type of media and the size
of the area to be scanned and displays an image on the screen.

The adjustment step comes next and involves cropping the image, choosing
a media type, selecting a scan resolution, and setting the exposure.
You can crop the image, selecting only the portion you want to copy, by
dragging the sides of the cropping outline the same way you adjust the
size of a window.

The media choices I have on my scanner are: line drawing, monochrome
photograph, color photograph, or millions-of-colors photograph.
Generally, I select the latter choice even when the scanner software
recommends another setting. The downside of doing this is it takes more
computer memory to store the image. The upside is more 'in-between'
information is detected. Examples of this in-between information are
places where the ink is faded and extremely faint or within a black
smear there is discernable writing that is just one shade blacker than
the rest. Selecting monochrome, or black and white, may result in the
loss of this subtle information. Later, if I deem the subtle
information unimportant and need to conserve the storage used, I may use
the graphics software to convert the image to monochrome.

The scan resolution I choose depends on the size of the original. Most
things we want to scan will be letter size or 8 1/2 x 11 inches. I
want to scan this to look good and be readable on the computer screen.
I've found that 100 dots per inch (DPI) works very well. I can just
barely discern the dots at 100 DPI. Lower resolutions appear jagged and
rough to me. If the image is smaller or if I want to magnify it, I
choose a greater resolution: 150, 200, or 300. The thing to remember is
the greater the resolution the more memory it takes to store the image.
An image, five inches square, scanned at 100 DPI takes at least 250,000
bytes of storage whereas the same image scanned at 300 DPI will take
2,250,000 bytes, minimum. I always use the same scan resolution value
for both horizontal and vertical settings to avoid distortion.

Once the above adjustments are made, it is time make the second scan.
The button in the scanner program window that does this may be labeled
ZOOM or RE-SCAN. Don't do another preview scan.

The last adjustment sets the exposure or lightness or darkness of the
image. This is done after the area to be copied is done in order to
fine tune the exposure. It is very important to do this step if the
area of interest is a different shade than the rest of the picture. I
usually use the automatic setting unless I can improve the readability
by doing it manually. My scanner program display has a button in the
middle of the window to do this. It is labeled with a circle that is
half dark. Other scanners do this automatically each time the image is
re-scanned. The scanner program may allow other adjustments, such as:
color tint, sharpness, contrast, and emphasis. Generally, these
adjustments are also offered by the graphics program as well. They are
too complicated to describe here. I've found the best way to learn
about them is by experimentation; trial and error. These additional
adjustments are useful for correcting the color of an old photograph (or
a poorly exposed new one) and for improving the readability of an old
document by emphasizing the middle shades while de-emphasizing the
darker
ones. If there is an interest, perhaps I can cover these techniques
another time.

After the exposure is adjusted, the output direction is selected. This
is usually done automatically by the software and you need not change it
unless you want to redirect the output to another place. I always go to
my graphics program because it has better manipulation, print and
storage options. I store all final pictures in the JPEG (*.jpg) format
which is the format endorsed by the Internet folks.

I will close by discussing more about resolution. When an image is
scanned, the hardware takes readings of the image at discrete points.
These are called dots. So 100 DPI resolution means that the image is
sampled, or read, at 100 points in each inch and 100x100 or 10,000
points in a square inch. If millions-of-colors or 32-bit color is
selected, for each dot the scanner selects which of 16.7 million colors
is closest to the one sampled and stores its value. If you get
impatient with the speed of your scanner, just think of all the
decisions it is making to digitize an image! At 100 DPI, a letter-sized
sheet is digitized to form an image that is 850 dots across horizontally
and 1100 dots down vertically.

A computer display screen displays dots per inch except on a display
they are called pixels. Today, most displays are at least 1024 pixels
across and 768 pixels down although sometimes the user will select and
use a lower resolution, either 800x600 or 640x480 pixels. Web browsers
and other programs map a scanned image to the screen, one dot for each
pixel, so a scanned image, 850x1100 pixels, fits nicely on a screen
although scrolling is sometimes required to see the bottom part of the
picture. That resolution also is sized such that it is easy to read.
If, however the image is small, say 2 inches square, I will enlarge it
to make it easier to see on the screen. A scanner resolution of 300
DPI on a two inch square original will yield a displayed image of 600
pixels by 600 pixels.

Ink Jet printers, the type most of us use, have the capability to print
300 dots per inch in color. Sometimes the resolution for black and
white printing is 600 DPI or greater. This maximum resolution is used
on the best setting with the best paper. Ordinary printing is usually
done at a lesser resolution. The mapping of image dots to printer dots
is not as exact as the mapping between images and the computer screen so
you should have at least twice as great printer resolution as available
in the image. An image using 100 dots per inch resolution will
usually print OK on most printers.

I've tried to give an quick overview of the scanning process without
getting bogged down with details. If anyone needs more or specific
help,
just send me a note.

Tony Cox
%%%%%%%%%%%%%
http://home.att.net/~ztlcox/
%%%%%%%%%%%%%

This thread:

• DUKE/FOSTER: Scanner How-to by "Tony Cox" <>