Path: utzoo!attcan!uunet!lll-winken!sol.ctr.columbia.edu!samsung!cs.utexas.edu!yale!bunker!wtm From: campbell%hpdmd48@hplabs.HP.COM (Gary Campbell) Newsgroups: misc.handicap Subject: Re: ARKENSTONE Response Message-ID: <15121@bunker.UUCP> Date: 24 Oct 90 03:28:31 GMT References: <15018@bunker.UUCP> Sender: wtm@bunker.UUCP Reply-To: campbell%hpdmd48@hplabs.HP.COM (Gary Campbell) Distribution: misc Lines: 63 Approved: wtm@bunker.UUCP Fidonet: Blink Talk Conference Index Number: 11244 Les Barr (Les.Barr@p0.f41.n147.z1.fidonet.org) writes: >As far as using the HP ScanJet and bound books (Hardback & >softback) the article really missed the mark. I scan almost I can't remember scanning a hardback book, but with paperbacks, I find that it takes some care to make sure that the page is straight on the scanner. I often have to keep my hand on the binding so that the pressure from the binding does not cause the page to skew during scanning. If it's my paperback, and I have the time to go get it done, I prefer to chop off the binding and throw it in the auto document feeder. Flattening the binding, to the point that it starts to break down a bit makes things work better. This is something I would not want to do with a library or borrowed book. While things that seem hard at first can get easier with practice and a little experimentation, I think that to say that Chip "really missed the mark" is a bit strong. We ScanJet users may not know how easy the Kurzweil users have it. >exclusively rather large, bound books. (For college work) And I >have had no difficulty in using my HP scanner with any book I've >read thus far. I am scanning/reading a book called, "Contemporary >Literary Criticism," and I have to write a graded journal entry I think it should be pointed out that there is a difference between literary books and technical books such as computer texts, computer reference books, and math texts. While I find TrueScan works quite well for literary text, I do not find it alone satisfactory for something for which I need to know exact syntax. It should be no surprise to anyone who knows anything about ASCII and print math symbols (I think some ancient mathematician dipped his favorite nightcrawler in an ink well and let it loose on a piece of paper ) that TrueScan won't scan them, but I find that it doesn't do well with many ASCII symbols in computer programs. For example, I often get braces {} or [] as parentheses, ~ ` and ^ are also likely to be misrecognized, and I have known times when some of them vanish altogether. I do find it useful to scan these kind of books because I have an optacon which I can use to get the exact syntax when I need it, but I would be very disappointed if I were expecting to be able to read a book on a new programming language complete with statement syntax, bought an Arkenstone, and then found out it isn't as accurate as I thought. I have heard that the KPR does the same kind of thing, but possibly on different characters, but have no experience or reliable reports to go on. On the whole I have been very happy with the TrueScan, and am considering buying one for home. >content. Works on the same principle as "scan now, translate >later." Wherein, you would scan the print into image files onto >your hard disk, and allow ARKENSTONE to translate the files while >you are doing something else. I generally use this feature when >reading 20 pages or more. This way, when I return, my >text/WP/WS/etc. file is ready to be used. Do you know what is the best image format for this? I find that TIFF packbits seems to scan about as fast as PCX and is smaller, and that TIFF compressed G3 or whatever it's called is smaller yet but takes about twice the time to scan. -- Gary Campbell Internet: campbell@hpdmd48.BOI.HP.COM