home *** CD-ROM | disk | FTP | other *** search
- Path: sparky!uunet!cs.utexas.edu!qt.cs.utexas.edu!yale.edu!spool.mu.edu!umn.edu!lynx!pprg.unm.edu!daemon
- From: Brian Bartholomew <bb@math.ufl.edu>
- Newsgroups: comp.soft-sys.khoros
- Subject: A series of simple-ish questions
- Message-ID: <42322@pprg.eece.unm.edu.pprg.unm.edu>
- Date: 24 Dec 92 03:10:39 GMT
- Sender: daemon@pprg.unm.edu
- Lines: 56
-
- (Good grief, I feel ignorant. Bear with me, please. I promise I'll
- take that Intro to Image Algebra course next semester.)
-
- I have a version of our University seal which I've helped to digitize
- into PostScript splines in a drawing package. It looks good printed 8
- inches across at 300 dpi.
-
- (a) I also want to print it 0.6 inches across at 300 dpi for
- letterhead. It looks cruddy that way, as many of the details that are
- outlined by splines are smaller than 1/300 of an inch; PostScript
- fills in too many dots due to its coloring rule.
-
- (b) If I render the PostScript into a bitmap and have PostScript
- shrink and print that, it looks much better, because PostScript is
- smart about such things. However, I think I can do better.
-
- (c) The result I'd rather have is one an artist trained in hand-tuning
- bitmap fonts produces; angled lines are suggested by customized
- halftones which serve to antialias them. Features smaller than 300
- dpi are "greeked", and so forth.
-
- I'd like to find out how far Khoros can take me to (c). I can use
- Ghostscript or my NeXT to produce a bitmap of the logo at a large
- size. I suspect I can then "blur" or "antialias" or "low-pass filter"
- that bitmap with Khoros [details?]. I suspect I can then "dither" or
- "halftone" this gray image into a B/W bitmap [details?] which would
- look better on my printer than (a) and (b) but not as good as (c).
-
- Am I correct to say:
-
- A "digital halftone" turns a single gray pixel into a regular
- arrangement of B/W pixels, which are set by a rule called a
- "spot function", which when low-pass filtered by the eye
- approximates the original gray.
-
- A "dither" turns a single gray pixel into a single B/W pixel,
- but the region around a B/W pixel approximates the value of
- region around the gray pixel when low-pass filtered by the
- eye. Dither algorithms are usually more clever than halftone
- algorithms, because the job is harder.
-
- You could combine the above two to make a halftone scheme with
- a clever spot function that eliminates bumps in the gamma
- curve and visual artifacts caused by regular shapes appearing
- in the halftone cells.
-
- Maybe the difference is best stated as "a halftone never looks
- at values outside the source gray pixel".
-
- I have Khoros v1.0. Thanks for your time and whatever light you can
- shed on this stuff, and Season's Greetings.
-
-
- Another member of the League for Programming Freedom (LPF) lpf@uunet.uu.net
- -------------------------------------------------------------------------------
- Brian Bartholomew - bb@math.ufl.edu - Univ. of Florida Dept. of Mathematics
-