Quote:
Originally Posted by Taylor514ce
Thanks, Hadrien. The Sony Reader links (at least on the one book I checked) give a PDF, not an LRF.
So using this particular author as an example, say I wanted to create ebook versions of his works. I'm looking for a general outline.
1. Research copyright. How? Where? Tools, databases of works?
2. Find content. What sites? Do any of you scan/OCR your content?
3. Format as ebook. I know there are tools/tutorials here, I can research that.
4. Upload
|
I've done all of it except for the copyright research, (I have done several keel-up scans of 19th century books (clearly P.D.) for P.G Australia). It's not simple and easy, but you get that "craftsman" sense of accomplishment when done (at least I did).
Equipment - OMNISCAN 3600 Scanner, Windows Professional SP2 computer, ABBYY Finescan OCR software (comes with the scanner). Budget 1-2 days per book.
After installing the scanner software and powering up the scanner, a typical book goes as follows.
I scan either 1 page per scan or 2 pages (flat) per scan. 1 page if the scanner won't hold the book flat. I scan 50 to 100 pages, the OCR software tends to choke on large scan files than that. Once the OCR is done, save each block of text with a different name. I save it in RTF format, as I find that format is the easiest to edit later. Repeat until the whole book is scanned.
Not the
real work begins... Load the first block into Wordpad. (I find Wordpad simplier to proofread with than Word.) Put the book in your lap, and start reading an OCR block. Anything that looks odd in the block, compare with the text, and fix as necessary. You can't afford to get lost in the story, you're working... After you have done all the OCR block, use 2 wordpad windows and cut and paste the blocks together. Once the book is together, reset the margins to a consistant margin. Save the resulting e-book. Then take it and start the proof job all over again (you
will find mistakes you missed the first time.) The OCR software is about 99.5 percent accurate, which means around 1 error per page....
By this time you will be heartily sick of the book you are working on. (after 2 pleasureless proofreading reads). Fire up Word and load the RTF book into Word. Save it as HTML. The result will be a single page HTML e-book (hurrah, send white smoke up the chimney!!) I don't know HTML, so I can't do the fancy hyperlinks. I also don't do LRF/MOBI/PDF, so for those you're on your own. (That's why I don't do uploads to Mobileread.)
Warning, the equipment isn't cheap, and it's
work. But to me, there's the satisfaction of adding something to the e-book world.
Pro Bono Publico
Your mileage may vary......
P.S. If you're going to do multiple formats, get a good file hex editor. You'll
need it...