Quote:
Originally Posted by lkmiller
You might actually have better luck with topaz with an older version of calibre and DeDRM. There may be some things that got lost in the switch from python 2 to python 3. But that might be more trouble than it's worth because it still may not work.
|
Co-incidently I was reading about Topaz a few days ago. One version was actually a cunning idea:
1) Scan paper copy
2) The software does the usual OCR
AND
3) The software pattern matches all the images of the letters and generates a kind of average custom glyph rescaleable "font".
4) The Topaz file is then reflowable copy of scanned paper as images are replaced by indexes to the glyph table and single newlines are replaced by space. The OCR text is used for search etc.
I've never seen one. But the glyphs would be usually unique per book. However a 167 dpi or 150 dpi rendering of the glyphs might have been rather rough compared to a proper font on the Viziplex screen.
It sounds like a nice idea in theory, but obviously actual proof reading and editing produces a better result. Unlike mobi originally (though azw is Amazon encryption) which could use mobipocket encryption, I think it was Amazon encryption. I'm not clear on differences between .tpz, azw1 and azw2 versions.
When did they abandon it? Apparently there was once a PC program to create topaz files from scans?
It would certainly be much smaller than a scan in a PDF container.