berger@clio.las.uiuc.EDU (06/01/88)
What does a 3% deviation in the resolution of a laser printer MEAN?
Mike Berger
Department of Statistics
Science, Technology, and Society
University of Illinois
berger@clio.las.uiuc.edu
{ihnp4 | convex | pur-ee}!uiucuxc!clio!bergerBNB@SEED.AMS.COM (b beeton) (06/22/88)
in practical terms, a 3% deviation in the resolution of a laser printer
means that if you expect your print area to be 10in long, it could be
as much as .3in (about 1.8pc) shorter or longer. if you're unlucky
enough to get opposite variations vertically and horizontally, a square
you expect to be 10pc on a side could actually end up being 10.3 x 9.7pc
and that's quite enough to be noticeable! it also means that if you're
trying to follow a designer's specs, that are ultimately targeted for
a typesetter, and you get complaints that your dimensions aren't precise
("the page is a pica too short"), you can't rely on laser proofs, but
must get typesetter output to verify the design implementation.
a test we've used to determine the deviation of the printers we're using
consists in drawing several parallel lines of specified lengths (we use
3, 4, 5 and 6in, drawn .5in apart) and printing the results once in
portrait and once in landscape mode. on one machine with a xerox xp-24
engine, the fixit person has been able to get the variation between
portrait and landscape down to about a point; on another, with a canon-cx
engine (with a 2% permissible deviation), the difference is about a pica
over the 6in rule, and we can't do anything about it. if precision is
very important to you, be warned -- read the fine print.
-- barbara beeton
american math society
-------