On 23/01/2008, martin f krafft <[EMAIL PROTECTED]> wrote: > also sprach Jeffrey Ratcliffe <[EMAIL PROTECTED]> [2008.01.16.2027 +0100]: > > A better option would have been to have saved the 170 pages as either > > a TIFF, which would be reimported as TIFF - better compression, or a > > DjVu file, which has anyway better compression. > > I tried both now. When trying to import the DjVu file, gscan2pdf > just says "Converting to TIFF", but the progress bar never fills. > > In the mean time, gscan2pdf inflicts some serious work on my hard > disks: > > it spawns ddjvu to convert my 12Mb .djvu file into a 230Mb TIFF, and > then apparently loads that multiple times, causing my entire 4Gb of > RAM and all of swap to fillto about 99%:
So the djvu->tiff process finishes, and it is the tiffsplit and import process that is producing the problems? > If I import the TIFF directly, then the same thing happens, except > for the .djvu stage. At the moment, as you have already seen, djvu import is: ddjvu to produce one (perhaps multipage) TIFF tiffsplit to split the TIFF to a single file per page import each page I'm thinking about a similar technique to that used for the PDF import - see how many pages the djvu has and import them individually: djvudump to dig out the number of pages run ddjvu once per page to dig out each page import each page Similarly, to import a multipage TIFF directly: tiffinfo to dig out the the number of pages run tiffcp once per page to dig out each page (not possible with tiffsplit) import each page In both cases, I'd display a dialog allowing the possibility to import a subset of pages. -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
