scanners, cont. (68)
Willard McCarty (MCCARTY@VM.EPAS.UTORONTO.CA)
Wed, 15 Feb 89 21:03:11 EST
Humanist Mailing List, Vol. 2, No. 611. Wednesday, 15 Feb 1989.
(1) Date: Tue, 14 Feb 89 18:35:16 PST (9 lines)
From: Jody Gilbert <USERDOG1@SFU.BITNET>
Subject: Scanning software for Macs (30)
(2) Date: Wed, 15 Feb 89 12:03:57 EST (39 lines)
From: Allen Renear <ALLEN@BROWNVM>
Subject: Omnipage
(1) --------------------------------------------------------------------
Date: Tue, 14 Feb 89 18:35:16 PST
From: Jody Gilbert <USERDOG1@SFU.BITNET>
Subject: Scanning software for Macs (30)
If we are talking about Optical Character Recognition software as
opposed to simple scanning software, I have also seen Read It (by
Olduvai I think) and TextPert. Read it worked OK, but the copy of
TextPert we got wouldn't open files that had been scanned by other programs,
although it was supposed to.
(2) --------------------------------------------------------------43----
Date: Wed, 15 Feb 89 12:03:57 EST
From: Allen Renear <ALLEN@BROWNVM>
Subject: Omnipage
Andrew Gilmartin, Brown CIS, forwards the following on Omnipage.
------------------------------------------------------------------------
I have been using OmniPage for a short while now mostly to sample the
range of original page quality it will except. In general OmniPage does
not like photocopies of articles typeset at below 12pt. For example, a
typical Byte magazine article which uses a serif font in a multicolumn
layout with headings crossing columns and interspersed graphics will be
read with 100% accuracy while a high quality photocopy of the same
article is unusable. Bold text is read accurately. Italic text is
problematic, the finer its strokes the poorer the result. I now simply
plan to retype the italic sections.
Earlier this week we tried scanning a photocopy of an Italian document
printed on a Macintosh ImageWriter with 12pt New York (fully
justified). More care must to be taken in placing each page on the
scanner because of the ImageWriter's poor resolution but once this was
discovered the result was quite acceptable. 'Acceptable' is a little
fuzzy so let me say that the scan resulted in a document that could be
given to a student to cleanup in a half an hour or so.
Others in our office have scanned everything from tables of data to the
front page to IBM documentation. Each time the resulting document would
have taken much longer to have had retyped than it took to scan and
clean. It has been my experience that a five to ten page typeset
article takes a little over an hour to scan, clean up, and reformat.
The equipment used is an Apple flatbed scanner connected to a Macintosh
II running OmniPage under MultiFinder with the default memory settings.
Andrew Gilmartin at Computing & Information Services
andrew@brownvm.brown.edu
ps OmniScan can distinguish between a hyphen and an em-dash!