| [Thread Prev] | [Thread Next] | [Thread Index] | [Date Prev] | [Date Next] | [Date Index] |
Terry S. Arnold wrote:
>A colleague of mine is in the hroes of fininshing a monograph on the radula
>of Cyraeidae. This monograph is illustrated with several hundred SEM and
>optical microphotographs. He is attempting to use scanned verions of the
>photographs in the electonic publication process which will result in an
>ultimate paper product. At the present time he is working with the printer
>ot come up with the right combination of scanning pitch (DPI) and file
>format that will produce the best possible result within a fairly limited
>budget. The attempts at 300 DPI appear to only be producing a result with a
>net resoltion of 85-87 DPI after being scanned, going through Photoshop for
>image cleanup, transferring in the form of TIFF files to the printer, and
>printing on a 600 DPI Docutek printer. On the other hand proofs printed on a
>color laser printer come out much better. The 85-87 DPI is not good enough
>for the subject matter which has significance in some of the fine detail.
>
>What suggestions or experiences do you have related to this type of problem.
>As more of us go toward electronic preparation of illustrations this
>situation will be faced more and more often.
There is some simple mathematics that can answer at least your basic
question. An ordinary TIFF file uses 8-bit pixels and can thus handle 255
(2^8-1, discounting the pure white and black) shades of grey. When you
print to paper with an offset printing press, laser printer, etc., you use
only black ink or powder, so the greys are formed by different sizes of
dots in halftone screens.
Thus there are two different kinds of resolution to keep in mind: One is
the resolution of your printer (i.e. how small black dots can it print),
expressed in DPI (dots per inch) and the other one is the resolution of
your halftone screen, expressed in LPI (lines per inch).
The printer forms the different sizes of halftone dots by building them up
from different numbers of black blocks (the printer's dots). In order to
represent 255 shades of grey, your printer then needs to print with a
resolution that is 16 times finer than that of the halftone screen (a
square with 16 dots along a side can represent 257 different values from 0
[white] through 1-255 [greys] to 256 [black]).
Whereas many pictures can get away with fewer greys than 255, it's not
generally recommendable. Particularly gradations over larger surfaces
quickly acquire a stepped appearance if there the number of grey levels is
insufficient.
As for the resolution of the halftone screen, this again is a matter of
choice, but for acceptable printing of fossil micrographs you should not go
below 130 LPI. Your printer should then have a resolution of about 2000
DPI. If you use a printer with 600 DPI you will have to sacrifice
resolution or greyscale, preferably both. If your net resolution of 85-87
DPI refers to the halftone screen's LPI using a 600 DPI printer, you
probably have only around 50 levels of grey preserved in the picture and
the halftone dots will be clearly visible to the eye.
There are some technical tricks some printers use to improve the
performance (for example the use of edge smoothing may make a difference),
but basically you need a printer with considerably higher resolution than
600 DPI in order to print decent halftones.
The good side of the story is that your scanned image doesn't have to have
a particularly high resolution. The recommended value is about 1.55-2 times
the resolution of the final halftone screen. Printing with a 130 LPI screen
would then call for a scanned resolution of about 200-260 DPI. Anything
finer than that would only increase file size without any appreciable
improvement of image quality.
Stefan Bengtson _/ _/ _/_/_/ _/ _/
Department of Palaeozoology _/_/ _/ _/ _/ _/_/ _/_/
Swedish Museum of Natural History _/ _/ _/ _/ _/ _/ _/ _/ _/
Box 50007 _/ _/ _/ _/_/_/ _/ _/ _/
S-104 05 Stockholm _/ _/_/ _/ _/ _/ _/
Sweden _/ _/ _/ _/ _/ _/
tel. +46-8 666 42 20
+46-18 54 99 06 (home)
fax +46-8 666 41 84
e-mail Stefan.Bengtson@nrm.se
Partial index: