Manuals to PDF advice sought

From: Marion Bates <Marion.Bates_at_dartmouth.edu>
Date: Sun Jun 10 11:13:37 2001

Hi,

I've run into the same difficulties trying to do the same thing before, using Acrobat 3. I found that if I used the actual Acrobat programs (duh can't remember the names of the individual "pieces") then I always wound up with huge files. BUT if I used the silly little AcrobatPDFWriter printer driver to "print" PDFs from Photoshop, they are always much smaller...undoubtedly this is due to my not understanding what the heck I'm doing with Acrobat Exchange (ah that's the name) but maybe give that a try if you haven't already.

Also, and this is something you probably already know, but unless the stuff you're scanning contains images with gradients (i.e. photographic) you can drop the bitlevel to 2 (true black-and-white) and up the scanning DPI to 300 or 600 and get very sharp results with that. They will print like new. If you then print _that_ from Photoshop to the PDFWriter you'll probably get some good results.

I think Acrobat has some compression options as well. Can't really remember the details.

Re: OCR -- Trouble I've had is (and this is just pickiness, if the actual info's all you care about then it's no prob) you invariably lose the font and other aspects of the original appearance of the document, which is a bummer. I converted a PDF of Sun Remarketing's Lisa DIY guide into HTML with images because I wanted search engines to be able to index the content. I used Omnipage Pro 8 which has a very convenient save-to-HTML feature and it even knows to pull out the images and save those as high-res JPEGs. So if you do end up going the OCR route, Omnipage is great. Textbridge Pro is "smart" about reading text accurately, but I think it lacks the features of Omnipage.

You could also send me a sample scanned image (try a few different resolution combinations) off-list and I'll see what kind of results I can get, and send 'em back...

HTH. Good luck!

-- MB

--- classiccmp_at_classiccmp.org wrote:
        I've decided that I would like to start converting various documentation I have to PDF, mainly newsletters and such for the Model 2000 that seem to be otherwise nearly impossible to find. In fact, I'd like to eventually cut a CD that pretty much covered the machine and it's docs. My question is, how do I best go about doing this? I've tried a couple of newsletters already and scanning them using Photoshop as a grayscale TIFF's at 100dpi, so as to preserve botht the layout and any diagrams and such, and then importing the images into Acrobat ended up with a 6.3meg PDF for a 20page 8.5x11 newsletter. If I scan the newsletters as a bi-tone image the text comes out all blotchy. I also scanned a 4 page 8.5x11 glossy color flyer, using truecolor settings at 200dpi, and it ended up as a 32meg file.

        I don't have any OCR software but I do have Acrobat 4.0 and Photoshop 2.5.1. Yes, I know it's an extremely old version of Photoshop but it does the job and Acrobat doesn't like my scanner plugin while Photoshop works with it fine. Acrobat PDF Writer works from other programs fine as well, having easily converted my FAQ to a 32k PDF by 'printing' to it from within Netscape. The scanner is a Piotech Splendeur 3024, a rebadged Relisys Infinity Scorpio, capapble of 30bit color at a maximum optical resolution of 300 x 1200. It's capable of 2400 x 2400 with software interpolation.

        Any suggestions would be greatly appreciated.

        Thanks
        Jeff
--- end of quote ---
Received on Sun Jun 10 2001 - 11:13:37 BST