How Computers Work: Input and Output
The central processing unit is the unseen part of a computer system, and
users are only dimly aware of it. But users are very much aware of the
input and output associated with the computer. They submit input data to
the computer to get processed information, the output.
Sometimes the output is an instant reaction to the input. Consider these
- Zebra-striped bar codes on supermarket items provide input that permits
instant retrieval of outputs - price and item name - right at the checkout counter.
- A bank teller queries the computer through the small terminal at the
window by giving a customer's account number as input. The same
screen immediately provides the customer's account balance as output.
- A forklift operator speaks directly to a computer through a microphone.
Words like left, right, and lift are the actual input data. The output is
the computer's instant response, which causes the forklift to operate as
- A medical student studies the human body on a computer screen,
inputting changes to the program to show a close-up of the leg and then
to remove layers of tissue to reveal the muscles and bone underneath.
The screen outputs the changes, allowing the student (without donning
a mask, sanitary gloves, or operating gown) to simulate surgery on the
- A sales representative uses an instrument that looks like a pen to enter
an order on a special pad. The handwritten characters are displayed as
"typed" text and are stored in the pad, which is actually a small computer.
Input and output may sometimes be separated by time or distance or
both. Here are some examples:
- Factory workers input data by punching in on a time clock as they go
from task to task. The time clock is connected to a computer. The outputs
are their weekly paychecks and reports for management that summarize
hours per project on a quarterly basis.
- A college student writes checks. The data on the checks is used as input
to the bank computer, which eventually processes the data to prepare a
bank statement once a month.
- Charge-card transactions in a retail store provide input data that is
processed monthly to produce customer bills.
- Water-sample data is collected at lake and river sites, keyed in at the
environmental agency office, and used to produce reports that show
patterns of water quality.
The examples in this section show the diversity of computer applications, but in all cases the process is the same: input-processing-output. We
have already had an introduction to processing. Now, in this chapter we
will examine input and output methods in detail.
Input: Getting Data from the User to the Computer
Some input data can go directly to the computer for processing. Input in
this category includes bar codes, speech that enters the computer through a
microphone, and data entered by means of a device that converts motions
to on-screen action. Some input data, however, goes through a good deal
of intermediate handling, such as when it is copied from a source document
and translated to a medium that
a machine can read, such as a magnetic disk. In either case the task is to
gather data to be processed by the computer ‹sometimes called raw
data and convert it into some form the computer can understand.
A keyboard is usually part of a personal
computer or part of a terminal that is connected to a computer
somewhere else. Not all keyboards are traditional, however.
A fast-food franchise like McDonald's, for example, uses keyboards whose
keys represent items such as large fries or a Big Mac. Even
less traditional in the United States are keyboards that are used to
enter Chinese characters.
A mouse is an input device with a ball on its underside that is rolled on
a flat surface, usually the desk on which the computer sits. The rolling
movement causes a corresponding movement on the screen. Moving the
mouse allows you to reposition the pointer, or cursor, an indicator on the
screen that shows where the next interaction with the computer can take
place. The cursor can also be moved by pressing various keyboard keys.
You can communicate commands to the computer by pressing a button on
top of the mouse. In particular, a mouse button is often used to click on an
icon, a pictorial symbol on a screen; the icon represents a
computer activity-a command to the computer-so clicking the icon
invokes the command.
A variation on the mouse is the trackball. You may have used a trackball
to play a video game. The trackball is like an upside-down mouse-you
roll the ball directly with your hand. The popularity of the trackball surged
with the advent of laptop computers, when traveling users found them-
selves without a flat surface on which to roll the traditional mouse.
Source Data Automation: Collecting Data Where It Starts
Efficient data input means reducing the number of intermediate steps
required between the origination of data and its processing. This is best
accomplished by source data automation ‹the use of special equipment to
collect data at the source, as a by-product of the activity that generates the
data, and send it directly to the computer. Recall, for example, the supermarket
bar code, which can be used to send data about the product
directly to the computer. Source data automation eliminates keying,
thereby reducing costs and opportunities for human-introduced mistakes.
Since data about a transaction is collected when and where the transaction
takes place, source data automation also improves the speed of the input
For convenience, we will divide this discussion into the primary areas
related to source data automation: magnetic-ink character recognition,
optical recognition, data collection devices, and even directly by your own
voice, finger, or eye. Let us consider each of these in turn.
Magnetic-Ink Character Recognition
Abbreviated MICR, magnetic-ink character recognition is a method of
machine-reading characters made of magnetized particles. The most common
example of magnetic characters is the array of numbers across the
bottom of your personal check.
Most magnetic-ink characters are preprinted on your check. If you
compare a check you wrote that has been cashed and cleared by the bank
with those that are still unused in your checkbook, you will note that the
amount of the cashed check has been reproduced in magnetic characters in
the lower-right corner. These characters were added by a person at the
bank by using a MICR inscriber.
An inexpensive way to get entire documents, pictures, and anything on a flat
Surface into a computer is by using a scanner. Scanners use
optical recognition systems that have a light beam to scan input data to convert
it into electrical signals, which are sent to the computer for processing.
Optical recognition is by far the most common type of source input,
appearing in a variety of ways: optical marks, optical characters, bar
codes, handwritten characters, and images. Scanners use Optical Character
Recognition software, described below, to translate text on scanned documents into text that is suitable for word processors and other computer applications.
Figure 1: Flatbed Scanner
Optical Mark Recognition
Abbreviated OMR, optical mark recognition is sometimes called mark
sensing, because a machine senses marks on a piece of paper. As a student,
you may immediately recognize this approach as the technique used to
score certain tests. Using a pencil, you make a mark in a specified box or
space that corresponds to what you think is the answer. The answer sheet
is then graded by a device that uses a light beam to recognize the marks
and convert them to computer-recognizable electrical signals.
Optical Character Recognition
Abbreviated OCR, optical character recognition devices also use a light
source to read special characters and convert them into electrical signals to
be sent to the central processing unit. The characters-letters, numbers,
and special symbols-can be read by both humans and machines. They are
often found on sales tags on store merchandise. A standard typeface for
optical characters, called OCR-A, has been established by the American
National Standards Institute.
The handheld wand reader is a popular input device for reading OCR-A.
There is an increasing use of wands in libraries, hospitals, and factories, as
well as in retail stores. In retail stores the wand reader is connected to a
point-of-sale (POS) terminal. This terminal is somewhat like a cash register,
but it performs many more functions. When a clerk passes the wand
reader over the price tag, the computer uses the input merchandise number
to retrieve a description (and possibly the price, if not on the tag) of the
item. A small printer produces a customer receipt that shows the item
description and price. The computer calculates the subtotal, the sales tax
(if any), and the total. This information is displayed on the screen and
printed on the receipt; notice that both screen and printer are output, so
the POS terminal is a complex machine that performs both input and output
functions. Finally, some POS terminals include a device that will accept
a credit card, inputting account data from the magnetic strip on a customer's
The raw purchase data becomes valuable information when it is
summarized by the computer system. This information can be used by the
accounting department to keep track of how much money is taken in each
day, by buyers to determine what merchandise should be reordered, and
by the marketing department to analyze the effectiveness of their ad
Each product on the store shelf has its own unique number, which is part
of the Universal Product Code (UPC). This code number is represented on
the product label by a pattern of vertical marks, or bars, called bar codes.
(UPC, by the way, is an agreed-upon standard within the supermarket
industry; other kinds of bar codes exist. You need only look as far as the
back cover of this book to see an example of another kind of bar code.)
These zebra stripes can be sensed and read by a bar code reader, a photo-
electric device that reads the code by means of reflected light. As with the
wand reader in a retail store, the bar code reader in a bookstore or grocery
store is part of a point-of-sale terminal. When you buy, say, a can of corn
at the supermarket, the checker moves it past the bar code reader.
The bar code merely identifies the product to the store's computer;
the code does not contain the price, which may vary. The price is stored in
a file that can be accessed by the computer. (Obviously, it is easier to
change the price once in the computer than to have to repeatedly restamp
the price on each can of corn. ) The computer automatically tells the point-
of-sale terminal what the price is; a printer prints the item description and
price on a paper tape for the customer. Some supermarkets are moving to
self-scanning, putting the bar code reader-as well as the bagging-in the
Although bar codes were once found primarily in supermarkets, there
are a variety of other interesting applications. Bar coding has been
described as an inexpensive and remarkably reliable way to get data into
a computer. It is no wonder that virtually every industry has found a niche
for bar codes. In Brisbane, Australia, bar codes help the Red Cross
manage their blood bank inventory. Also consider the case
of Federal Express. The management attributes a large part of the
corporation's success to the bar-coding system it uses to track packages. Each
package is uniquely identified by a ten-digit bar code, which is input to
the computer at each point as the package travels through the system. An
employee can use a computer terminal to query the location of a given
shipment at any time; the sender can request a status report on a package
and receive a response within 30 minutes. The figures are impressive: In
regard to controlling packages, the company has an accuracy rate of better
than 99 percent.
Machines that can read handwritten characters are yet another means of
reducing the number of intermediate steps between capturing data and
processing it. In many instances it is preferable to write the data and
immediately have it usable for processing rather than having data entry
operators key it in later. However, not just any kind of handwritten scrawl will
do; the rules as to the size, completeness, and legibility of the handwriting
are fairly rigid.
In a process called imaging, a scanner converts a drawing, a picture, or any
document into computer-recognizable form by shining a light on the image
and sensing the intensity of the reflection at each point of the image. Scanners
come in both handheld and desktop models. The electronic
version of the image can then be stored, probably on disk, and
reproduced on screen when needed. Businesses find imaging particularly
useful for documents, since they can view an exact replica of the original
document at any time. If a text image is run through an optical character
recognition (OCR) program, then all words and numbers can be manipulated
by word processing and other software. The Internal Revenue Service,
using imaging and also OCR software that can recognize characters
from the image, is now scanning 17,000 tax returns per hour, a significant
improvement over hand processing.
Another way to keep photos computer accessible is to have film that
was shot with a conventional camera processed onto optical disk instead
of prints or slides. Professional photo agencies keep thousands of images
on file, ready to be leased for a fee. Typically, a couple of
dozen thumbnail-size images can be displayed on the screen at one time; a
particular image can be enlarged to full-screen size with a click of a mouse
Data Collection Devices
Another source of direct data entry is a data collection device, which may
be located in a warehouse or factory or wherever the activity that is
generating the data is located. As we noted earlier in the chapter,
for example, factory employees can use a plastic card to punch job data
directly into a computerized time clock. This process eliminates intermediate
steps and ensures that the data will be more accurate.
Data collection devices must be sturdy, trouble-free, and easy to use
because they are often located in dusty, humid, or hot or cold locations.
They are used by people such as warehouse workers, packers, forklift
operators, and others whose primary work is not clerical. Examples of
remote data collection devices are machines for taking inventory, reading
shipping labels, and recording job costs.
Does your computer have ears? Speaking to a computer, known as voice
input or speech recognition, is another form of source input. Speech
recognition devices accept the spoken word through a microphone and convert
it into binary code (0s and 1s) that can be understood by the computer. Originally, typical users were those with "busy hands," or
hands too dirty for the keyboard, or with no access to a keyboard. Such
uses are changing radio frequencies in airplane cockpits, controlling inventory
in an auto junkyard, reporting analysis of pathology slides viewed
under a microscope, asking for stock-market quotations over the phone,
inspecting items moving along an assembly line, and allowing physically
disabled users to issue commands.
Most speech recognition systems are speaker-dependent; that is, they
must be separately trained for each individual user. The speech recognition
system "learns" the voice of the user, who speaks isolated words repeatedly.
The voiced words the system "knows" are then recognizable in the
Speech recognition systems that are limited to isolated words are called
discrete word systems, and users must pause between words. Experts have
tagged speech recognition as one of the most difficult things for a computer
to do. Eventually, continuous word systems will be able to interpret
sustained speech, so users can speak normally; so far, such systems are limited
by vocabulary to a single subject, such as insurance or the weather. A
key advantage of delivering input to a computer in a normal speaking pattern
is ease of use. Such systems may also be propelled by the explosion of
hand and wrist ailments associated with extensive computer keying.
Today, software is available to let computers take dictation from people
who are willing to pause . . . briefly . . . between . . . words; the best systems
are quite accurate and equivalent to typing 70 words per minute.
One way of getting input directly from the source is to have a human
simply point to a selection. The edges of the monitor of a touch screen
emit horizontal and vertical beams of light that criss-cross the screen.
When a finger touches the screen, the interrupted light beams can pinpoint
the location selected on the screen. Kiosks in public places such as malls
offer a variety of services via touch screens. An insurance
company kiosk will let you select a policy or a government kiosk will let
you order a copy of your birth certificate. Kiosks are also found in private
stores. Wal-Mart, for example, uses a kiosk to let customers find needed
auto parts. Many delicatessens let you point to salami on rye, among the
Delivering input to a computer by simply looking at the computer would
seem to be the ultimate in capturing data at the source. The principles are
reminiscent of making a screen selection by touching the screen with the finger.
Electrodes attached to the skin around the eyes respond to movement
of the eye muscles, which produce tiny electric signals when they contract.
The signals are read by the computer system, which determines the location
on the screen where the user is looking.
Such a system is not yet the mainstream. The first people to benefit
would likely be those who, due to disabilities or busyness, cannot use their
hands or voices for input.
Output: Information for the User
As we have seen, computer output takes the form of screen or printer
output. Other forms of output include voice, microfilm, and various forms
of graphics output.
A computer system often is designed to produce several kinds of output.
An example is a travel agency that uses a computer system. If a customer
asks about airline connections to Toronto, Calgary, and Vancouver, say,
the travel agent will probably make a few queries to the system to receive
on-screen output indicating availability on the various flights. After the
reservations have been confirmed, the agent can ask for printed output
that includes the tickets, the traveler's itinerary, and the invoice. The
agency may also keep the customer records on microfilm. In addition, agency
management may periodically receive printed reports and charts, such as
monthly summaries of sales figures or pie charts of regional costs. We
begin with the most common form of output, computer screens.
Computer Screen Technology
A user's first interaction with a computer screen may be the screen
response to the user's input. When data is entered, it appears on the screen.
Furthermore, the computer response to that data-the output-also
appears on the screen. Computer screens come in many varieties, but the
most common kind is the cathode ray tube (CRT). Most CRT screens use
a technology called raster-scan technology. The backing of the screen display
has a phosphorous coating, which will glow whenever it is hit by a
beam of electrons. But the light does not stay lit very long, so the image
must be refreshed often. If the screen is not refreshed often enough, the
fading screen image appears to flicker. A scan rate-the number of times
the screen is refreshed-of 60 times per second is usually adequate to
retain a clear screen image. As the user, you tell the computer what image
you want on the screen, by typing, say, the letter M, and the computer
sends the appropriate image to be beamed on the screen. This is essentially
the same process used to produce television images.
A computer display screen that can be used for graphics is divided into
dots that are called addressable, because they can be addressed individually
by the graphics software. Each dot can be illuminated individually on
the screen. Each dot is potentially a picture element, or pixel. The resolution
of the screen, its clarity, is directly related to the number of pixels
on the screen: The more pixels, the higher the resolution. Some computers
come with built-in graphics capability. Others need a device, called a
graphics card or graphics adapter board, that has to be added.
There have been several color screen standards, relating particularly to
resolution. The first color display was CGA (color graphics adapter),
which had low resolution by today's standards (320x200 pixels). This was
followed by the sharper EGA (enhanced graphics adapter), featuring
640x350 pixels. Today, VGA and SVGA are common standards. VGA
(video graphics array) has 640x480 pixels. SVGA (super VGA) offers
800x600 pixels or 1024x768 pixels, by far the superior clarity.
Is bigger really better? Screen sizes are measured diagonally. Many personal computers come with a 15 inch screen. A 15 inch screen is fine for most single applications, but for applications with large graphics, or for having multiple windows open, it is sometimes inadequate. For a few hundred dollars more, 17 inch can be better. There are even bigger screens that cost substantially more. Bigger is usually better, but more expensive.
Types of Screens
Cathode ray tube monitors that display text and graphics are in common
use today. Although most CRTs are color, some screens are
monochrome, meaning only one color, usually green, appears on a dark
background. Another type of screen technology is the liquid
crystal display (LCD), a flat display often seen on watches and calculators.
LCD screens are used on laptop computers. Some LCDs are
monochrome, but color screens are popular. Some laptop screens are nearing
CRTs in resolution quality.
A screen may be the monitor of a self-contained personal computer, or it
may be part of a terminal that is one of many terminals attached to a large
computer. A terminal consists of an input device, an output device, and a
communications link to the main computer. Most commonly, a terminal
has a keyboard for an input device and a screen for an output device,
although there are many variations on this theme.
A printer is a device that produces printed paper output, known in the
computer industry as hard copy because it is tangible and permanent
(unlike soft copy, which is displayed on a screen). Some printers produce
only letters and numbers, whereas others can also produce graphics.
Letters and numbers are formed by a printer either as solid characters
or as dot-matrix characters. Dot-matrix printers create characters in the
same way that individual lights in a pattern spell out words on a basketball
scoreboard. Dot-matrix printers construct a character by activating a
matrix of pins that produce the shape of the character. A traditional matrix is 5x7-that is, five dots wide and
seven dots high. These printers are sometimes called 9-pin printers,
because they have two extra vertical dots for descenders on the lowercase
letters g, j, p, and y. The 24-pin dot-matrix printer, which uses a series of
overlapping dots, dominates the dot-matrix market. The more dots, the
better the quality of the character produced. Some dot-matrix printers can
produce color images.
There are two ways of printing an image on paper: the impact method
and the non-impact method. Let us take a closer look at the difference.
The term impact refers to the fact that impact printers use some sort of
physical contact with the paper to produce an image, physically striking
paper, ribbon, and print hammer together. The impact may be produced
by a print hammer character, like that of a typewriter key striking a ribbon
against the paper, or by a print hammer hitting paper and ribbon against a
character. A dot-matrix printer is one example of an impact printer. High-
quality impact printers print only one character at a time.
However, users who are more concerned about high volume than high
quality usually use line printers - impact printers that print an entire line
at a time. Organizations that use mainframe and minicomputers usually
have several line printers. Such organizations are likely to print hearty
reports, perhaps relating to payroll or costs, for internal use. The volume
of the report and the fact that it will not be seen by customers makes the
speedy-and less expensive line printer appropriate. One final note
about impact printers: An impact printer must be used if printing a
multiple-copy report so that the duplicate copies will receive the imprint.
A non-impact printer places an image on a page without physically touching
the page. The major technologies competing in the non-impact market
are laser and ink-jet. Laser printers use a light beam to help transfer images
to paper, producing extremely high-quality results. Laser
printers print a page at a time at impressive speeds. Large organizations
use laser printers to produce high-volume customer-oriented reports. At
the personal computer end, low-end black and white laser printers can
now be purchased for a few hundred dollars. However, color laser jet
printers are more expensive.
The rush to laser printers has been influenced by the trend toward desktop
publishing-using a personal computer, a laser printer, and special
software to make professional-looking publications, such as newsletters.
Ink-jet printers, by spraying ink from multiple jet nozzles, can print
both black and white and in several different colors of ink to produce
excellent graphics. As good as they are, color printers are not perfect. The
color you see on your computer screen is not necessarily the color you will
see on the printed output. Nor is it likely to be the color you would see on
a four-color offset printing press. Nevertheless, with low-end printers now
under $250, they may be a bargain for users who want their own color
There are many advantages to non-impact printers over impact ones,
but there are two major reasons for their growing popularity: They are
faster and quieter. Other advantages of non-impact printers over conventional
mechanical printers are their ability to change typefaces automatically
and their ability to produce high-quality graphics.
We have already examined voice input in some detail. As you will see in
this section, however, computers are frequently like people in the sense
that they find it easier to talk than to listen. Speech synthesis is the process
of enabling machines to talk to people is much easier than speech recognition.
"The key is in the ignition," your car says to you as you open the
car door to get out. Machine voices are not real human voices. They are
the product of voice synthesizers (also called voice-output devices or
audio-response units), which convert data in main storage to vocalized
sounds understandable to humans.
There are two basic approaches to getting a computer to talk. The first
is synthesis by analysis, in which the device analyzes the input of an actual
human voice speaking words, stores and processes the spoken sounds, and
reproduces them as needed. The process of storing words is similar to the
digitizing process we discussed earlier when considering voice input. In
essence, synthesis by analysis uses the computer as a digital tape recorder.
The second approach to synthesizing speech is synthesis by rule, in
which the device applies a complex set of linguistic rules to create artificial
speech. Synthesis based on the human voice has the advantage of sounding
more natural, but it is limited to the number of words stored in the computer.
Voice output has become common in such places as airline and bus terminals,
banks, and brokerage houses. It is typically used when an inquiry
is followed by a short reply (such as a bank balance or flight time). Many
businesses have found other creative uses for voice output as it applies to
the telephone. Automatic telephone voices ("Hello, this is a computer
speaking. . . " ) take surveys, inform customers that catalog orders are ready
to be picked up, and, perhaps, remind consumers that they have not paid
Personal computer users have occasionally sent primitive musical messages,
feeble tones that wheezed from the tiny internal speaker. Many users
remain at this level, but a significant change is in progress.
Professional musicians lead the way, using special sound chips that simulate
different instruments. A sound card, installed internally in the computer,
and attached speakers complete the output environment. Now,
using appropriate software, the computer can produce the sound of an
orchestra or a rock band. Those of us who simply enjoy music can have a
full sight/sound experience using multimedia, which we will explore in
detail in the next chapter.
Let us take a moment to glimpse everyone's favorite, computer graphics.
Just about everyone has seen TV commercials or movies that use
computer-produced animated graphics. Computer graphics can also be found
in education, computer art, science, sports, and more. But
perhaps their most prevalent use today is in business.
It might seem wasteful to use color graphics to display what could more
inexpensively be shown to managers as numbers in standard computer
printouts. However, colorful graphics, maps, and charts can help managers
compare data more easily, spot trends, and make decisions more
quickly. Also, the use of color helps people get the picture-literally.
Finally, although color graphs and charts have been used in business for
years-usually to make presentations to higher management or outside
clients-the computer allows them to be rendered quickly, before information
becomes outdated. One user refers to business graphics as "computer-
Video graphics can be as creative as an animated cartoon.
Although they operate on the same principle as a moving picture or cartoon-one
frame at a time in quick succession video graphics are produced by
computers. Video graphics have made their biggest splash on
television, but many people do not realize they are watching a computer at
work. The next time you watch television, skip the trip to the kitchen and
pay special attention to the commercials. Unless there is a live human in
the advertisement, there is a good chance that the moving objects you see,
such as floating cars and bobbing electric razors, are computer output.
Another fertile ground for video graphics is a television network's logo and
theme. Accompanied by music and swooshing sounds, the network symbol
spins and cavorts and turns itself inside out, all with the finesse that
only a computer could supply.
Computer-Aided Design/Computer-Aided Manufacturing
For more than a decade, computer graphics have also been part and parcel
of a field known by the abbreviation CAD/CAM-short for computer-
aided design/computer-aided manufacturing. In this area computers are
used to create two- and three-dimensional pictures of everything from
hand tools to tractors. CAD/CAM provides a bridge between design (planning what a product will be) and manufacturing (actually making the
planned product). As a manager at Chrysler said, "Many companies have
design data and manufacturing data, and the two are never the same. At
Chrysler, we have only one set of data that everyone dips into." Keeping
data in one place, of course, makes changes easier and encourages consistency.
Graphics Input Devices
There are many ways to produce and interact with screen graphics. We
have already described the mouse; the following are some other common
devices that allow the user to interact with screen graphics. A digitizing
tablet lets you create your own images. This device has a special
stylus that you can use to draw or trace images, which are then converted
to digital data that can be processed by the computer.
For direct interaction with your computer screen, the light pen is ideal.
It is versatile enough to modify screen graphics or make a menu selection-that
is, to choose from a list of activity choices on the screen. A light
pen has a light-sensitive cell at one end. When you place the light pen
against the screen, it closes a photoelectric circuit that pinpoints the spot
the pen is touching. This tells the computer where to enter or modify pictures
or data on the screen.
Finally, a well-known graphics input device is the joystick, dear to the
hearts of video game fans. This device allows fingertip control of figures on
a CRT screen.
Graphics Output Devices
Just as there are many different ways to input graphics to the computer,
there are many different ways to output graphics. Graphics are most commonly
output on a screen or printed paper, as previously discussed.
Another popular graphics output device is the plotter, which can draw
hard-copy graphics output in the form of maps, bar charts, engineering
drawings, and even two- or three-dimensional illustrations.
Plotters often come with a set of four pens in four different colors. Most
plotters also offer shading features.
New forms of computer input and output are announced regularly, often
with promises of multiple benefits and new ease of use. Part of the excitement
of the computer world is that these promises are usually kept, and
users reap the benefits directly. Input and output just keep getting better.