General
Glossary of Image & Printing Terminology
Image Scanning &
resolutions:
- Determining
desired image resolution for output to computer screen or
video:
- Note if the image
is to be used solely for viewing on screen, then
no need for final image resolution to exceed the
screen resolution = screen pixels width / screen
inches width
- ie. on
15" monitor with 28cm/11"
displayable width then:
- 800x600
video resol. = 800/11 = 73dpi
screen resolution
- 640x480
on same monitor has resolution
of: 58dpi
- 1024x800
on same monitor has resolution of
93dpi
- For Full Screen
output:
- scanning
resolution = screen resolution width
(pixels) / original image width (inches)
- eg.
4"x3" image to 640x480
size then scan at 160dpi (640/4)
- For Actual Size
output on screen:
- res. =
screen resol. width (pixels) / max.
displayable screen width (inches)
- eg.
4"x3" image to appear
on 800x600pixel 11.75" wide
screen at 4"x3" image
then scan @ 68dpi
(800/11.75")
- For Scanning 35mm
slide/neg. (36x24mm) to size 100x67mm on 800x640
screen then scan at (100/36)*73dpi = 203dpi
- Determining image
resolution for output to printed media:
- desired scan res.
= (final image ht/orig. image ht) x halftone freq
(lpi) x 2
- eg.
4"x5" image to output as
3"x3.75" on laser printer with
60lpi then scan at 90dpi ((3.75/5)x60x2)
- Computer image
file sizes (uncompressed):
- = image height x
image width x (scan res.)^2 x bytes/pixel
Image type |
Bytes per pixel |
Bits per pixel |
No colors(2bits) |
Kbytes for 800x600
image |
B&W |
0.125 |
1 |
2 |
60 |
Indexed 16 color |
0.5 |
4 |
16 |
240 |
Indexed 256 color |
1 |
8 |
256 |
480 |
Grayscale |
1 |
8 |
256 |
480 |
16bit |
2 |
16 |
65,536 |
960 |
24bit "True color" |
3 |
24 |
16,777,216 |
1440 |
30bit CYMK |
3.75 |
30 |
1 x 109 |
1800 |
- Moire patterns:
- repetitive
patterns on scanned images esp. if scanned from
halftones
- minimise by
either:
- use
continuous tone image source
- scan at
twice resolution needed, then resample to
desired resolution
- try using
"Blur" or "Average"
to soften image
- if
speckled then try using
"Despeckle" function
Printing
Terminology:
- Print types:
- Continuous-tone
art: original photographs
- simulated
by: dye sublimation printers
- commercial
CYMK / black ink printing
- Halftone screens:
commercial printing using series of dots to make
a cell:
- "photographic
halftone" - dots can vary infinitely
in size
- "digital
halftone" dots fixed size
- CMYK
cyan/magenta/yellow/black - 4-color ink
printing
- screen or
halftone frequency or line frequency:
- # cells
per inch = lines per inch (lpi)
- screen or
halftone angle:
- orientation
of the line of halftone cells:
- B&W: usually
= 45deg.
- CYMK: black
45deg; M 75; Y 90; C 105
deg.
- dot gain:
- used to compensate
for spread of dot size due to absorbent paper or
thin ink used.
- printer resolution
(dots per inch / dpi):
- dot matrix 180
dpi;
- laser printer
300dpi;
- image setters 1270
- 2540 dpi;
- dpi/lpi =
#dots/cell
- paper types:
- weight
- absorbency
- eg.:
- newsprint:
<= 85lpi standard
- uncoated
stock: <=120lpi
- coated
stock: <= 300lpi
- Typesetting
measurements:
- pica typsetting
measurement approx. 1/6th of an inch = 12 points
- point typesetting
measurement = 1/12th of pica ie. approx. 72pts
per inch
- em typesetting
measurement = width of point size being used
- en typesetting
measurement = 1/2 em
- Pixel
"Picture Element"
- eg. VGA screen
640x480 pixels
- Digital computer RGB
format:
- 24bit color =
8bits per primary color (R/G/B) eg
640x480x24bits=921.6Kb;
-
- Image compression:
- Bezier drawing:
- similar to
"connect-the-dots" drawing
& can "pull" line to create
curves;
- Huffman coding -
compression technique
- Lempel-Ziv Welch
(LZW) compression technique
- DCT compression -
- processes
8x8pixel blocks
- no good
for dithered or quantised pictures as too
many high frequencies;
- YUV color
reduction - can reduce image file size by 30-60%
Image
Formats:
- Bitmap Images:
- BMP - MS Windows
bitmap format
- PCX - mainly used by paint
programs
- TGA - high quality bitmaps
in 16, 24 & 32bit formats - ?needs TARGA
video board;
- TIFF - various formats:
- GIF Compuserve compressed
images for modem data transfers;
- lossless -
uses Lempel-Ziv Welch (LZW) compression;
- SCD - SCODL created
by film recorders for slide making ? bitmap;
- RLE - run length
encoding compressed bitmap file - only good for
animation images;
- - rows of
adjacent identical pixels grouped;
lossless;
- JPEG Joint Photographers
Experts Group file format using superior
compression;
- various
formats: JFF, JTF, JPG, CMP (Lead
bitmaps));
- lossless -
uses only predictive & statistical
techniques;
- lossy
divides picture into 8x8 pixel blocks
& uses properties of DCT to reduce
data
- controlled
by quality factor Q which is a % &
controls quantification process prior to
Huffman coding - the lower Q the poorer
quality;
- Vector Graphics ie
"object-based" images:
- can be
converted to bitmaps by a "rasteriser"
- WMF - MS Windows
metafile
- CGM - Computer
Graphics Metafile
- CDR - created by
CorelDraw!
- WPG - Word Perfect 5
graphic
- PIC - 2 different
formats:
- Lotus
graphic
- created
by various slide-making
equipment;
- PIF - intermediate
format used to enable translation by IBM
mainframe to its GDF
- PLT plotter format for
drawings based on HPGL (HP Graphics Language);
- PCT - Macintosh
format - PICT 1 (B/W); PICT 2 (color);
- EPS - Encapsulated
PostScript file format for PostScript printers
- GEM - ? vector; created by
Graphics Environment Manager eg. Xerox Ventura
Publisher