Last updated: 2/3/2000
Link back to course Welcome

Forms of Computer Information

Computers can store and process many types of information. The different types of computer information are usually each stored in its own type of file, although some files can contain a mix of information types.

1. Program files - loaded into RAM and executed. Contents are computer instructions
   1. Operating system. The operating system provides common services such as organizing, reading and saving files, printing, displaying text and graphics
   2. Applications - specific functions such as word processing or graphics
2. Text. One character per byte, a numerical code for each character. Special characters in the top half of the set can be used for European characters, or for graphics characters, or for formatting codes in word processors and other applications
   1. Hypertext Markup Language files use only the normal characters in the bottom half of the character set. Formatting codes are enclosed in <> characters
   2. System text data such as passwords, default directories
   3. Codes for characters
      ASCII/ANSI CHARACTER CODES (Most codes less than 32 are not shown.
      Codes from 127 to 255 vary from cade to code and are not shown.)
      0 Null (nothing)
      7 Bell
      8 Back space
      9 Tab
      10 Line feed (new line at present column)
      13 Carriage return (return to beginning of line)
      26 End of file
      27 [Esc] (Escape key)
      32 [space]   64 @          96 `
      33 !         65 A           97 a
      34 "         66 B           98 b
      35 #         67 C           99 c
      36 $         68 D          100 d
      37 %         69 E          101 e
      38 &         70 F          102 f
      39 '         71 G          103 g
      40 (         72 H          104 h
      41 )         73 I          105 i
      42 *         74 J          106 j
      43 +         75 K          107 k
      44 ,         76 L          108 l
      45 -         77 M          109 m
      46 .         78 N          110 n
      47 /         79 O          111 o
      48 0         80 P          112 p
      49 1         81 Q          113 q
      50 2         82 R          114 r
      51 3         83 S          115 s
      52 4         84 T          116 t
      53 5         85 U          117 u
      54 6         86 V          118 v
      55 7         87 W          119 w
      56 8         88 X          120 x
      57 9         89 Y          121 y
      58 :         90 Z          122 z
      59 ;         91 [          123 {
      60 <         92 \          124 |
      61 =         93 ]          125 }
      62 >         94 ^          126 ~
      63 ?         95 _
   4. File size is determined by the number of characters, at one byte per character (plus some additional for formatting). For example, a page of simple text is 60 characters wide by 66 lines (about) or 3960 characters = 3960 bytes; about 4 kB per page.
   5. Text codes do not work for numerical computations. For example, using the codes, adding 1 and 2 would be 49 + 50 = 99, which is the code for the lower-case letter c. If numbers are represented as text codes, to do calculations they must first be converted to numbers. Fortunately, computers usually take care of this for us automatically.
3. Numerical data ("data")
   1. Dates, file size limits are in numerical form. Spreadsheets mix forms, with numbers being stored in numerical form, and labels being stored as text.
   2. Numbers are converted from/to text form for communication with the user
   3. File size is determined by the quantity of numbers, the accuracy with which the number is stored, and the way the number is represented. Integers can take from two to four bytes each, and floating point numbers (with a decimal point or an exponent) from four to eight bytes each.
4. Computers represent sound by converting sound waves to electrical voltages using a microphone, then measuring the voltage on a fast time scale and each time storing the number of volts (a "sample") as a number. This is also the way that sound is stored on a music CD, as a series of numbers. In the graphic below, the voltage samples are represented by the open squares.
   
   The file size is determined by the length of the sound, the number of voltages stored per second (sampling rate) and the accuracy with which each voltage is stored (one to two bytes each -- eight-bit or sixteen-bit sound).  The sound that plays when you start Windows 95 is The Microsoft Sound.wav. It has 22,050 one-byte samples per second and is a little over six seconds long, so the file size is 6 × 22,050 = 133 kB. This can be read quickly off your hard drive, but over the Internet via a modem, you would definitely want the file to be compressed.
   1. Normally, sound files are loaded completely into memory and then played. Over the Internet, this can take a long time, so "streaming" audio, which plays the sound as it is received, has become popular.
5. There are two ways of representing graphics or images in computers:
   1. "Bitmaps" divide the picture up into little squares, or "pixels".The color for each pixel is coded numerically. Since computer monitors generate colors by superimposing the right amount of red, green and blue, there is one number for the amount or intensity of red, another for green and a third for blue (the RGB scale). Here is the bitmap for the Word icon, shown normal size, and then the upper left part of the icon expanded about ten times so that you can see the pixels.
      

      Word Icon at left

      Upper left part of word icon, expanded to show pixels

      1. The most straightforward scheme is one byte or eight bits for each. This is "24 bit color", which can represent 256 × 256 × 256 = 16.8 million different colors. With this many colors, if the pixels are small enough (say 300 per inch, the standard for laser printers), the result is as good or better than a photograph, so this is sometimes called TrueColor. This scheme results in large files; a standard snapshot would be a 2.7 MB file.
      2. If color is not important, "grayscale" can give good results with one byte per pixel for 255 shades of grey. This is also as good as a "black and white" photograph.
      3. In a photograph, there are large areas where adjacent pixels have almost the same color, say as a surface goes from direct light to deep shadow, so the gradient or rate of change of the color can be used instead, allowing significant compression. This is the scheme used by JPEG (Joint Photographic Experts Group -- jpeg or jpg file format).
      4. We can also use fewer than one byte per pixel, with degraded quality.
      5. For graphics like posters, with not much shading, and large areas of the same color, and not too many colors overall. a "palette" of colors can be defined, with each color being given a numerical code. Typically, the palette is defined at the beginning of the graphics file, with the code and then the three RGB bytes for that color. Then the second part of the file has the color code for each pixel. A true black and white graphic need take only one bit per pixel. This is the scheme used by the GIF (Graphics Interchange Format) format, along with a further compression scheme.
   2. The second major scheme is to describe a picture as made up of things or objects, and to give the dimensions and color of each object. This is "vector" graphics, because the objects are positioned by vectors from a reference point. There are countless schemes here. One advantage is that the user can drag an object from one place to another, or delete it, or copy it. To be represented on a computer monitor or most printers, which are bitmap devices, a vector image must be converted first to bitmap form. For this reason, bitmap programs usually result in somewhat better graphics quality, but with a lot more work. Converting from vector to bitmap formats is standard, but it is very difficult (today) to convert a bitmap back into a vector format. Not to mention that it doesn't work very well.
      1. "Rendering" programs use the fastest computers to render a vector representation by tracing virtual light rays. Computer movies are made by having the computer move objects virtually, and then rendering each frame.
      2. "Virtual Reality" is lower in quality (today) but uses the same general scheme of moving objects and then drawing the picture, but in real-time.
   3. Programs that work with bitmaps generally have "paint" or "photo" in their name (Paint, Paintbrush, Paintshop, Photoshop) while programs that work with vector formats generally have "draw" in their name (Draw).
   4. Windows icons use a palette of sixteen colors and are 32 pixels on a side; the file extension is ico. Bitmap files have the extension .bmp. The Windows Metafile format (wmf) can mix bitmap and graphics.
   5. Computers represent motion, as in 2a above, by redrawing the screen rapidly. The MPEG (Motion Picture Experts Group) file format (mpg) is a standard. It achieves compression by only storing the changes between frames. Other formats are Apple's QuickTime and Microsoft's Audio Visual Interleave (avi).