> > If you compare your output with the original, you'll notice that the > "pdftops -level3 -paper match" > part is missing from your command. Also not that you should replace $infile > and $outfile with the > actual filenames you use (or set them as shell variables). > > Sample: > > infile=myoriginalpdf ; outfile=myconverted.pdf; pdftops -level3 -paper match > "$infile" - | gsc -q -dSAFER -dBATCH -dNOPAUSE -sDEVICE=pdfwrite > -dPDFSETTINGS=/ebook -dAutoRotatePages=/None -dEmbedAllFonts=true > -dSubsetFonts=true -sOutputFile="$outfile" -dCompatibilityLevel=1.4 -c > .setpdfwrite - > > Or, create a shellscript for it: > __________ > #!/bin/sh > infile=$1 > outfile=$2 > > pdftops -level3 -paper match "$infile" - | gsc -q -dSAFER -dBATCH -dNOPAUSE > -sDEVICE=pdfwrite -dPDFSETTINGS=/ebook -dAutoRotatePages=/None > -dEmbedAllFonts=true -dSubsetFonts=true -sOutputFile="$outfile" > -dCompatibilityLevel=1.4 -c .setpdfwrite - > __________ > > > Put that in a file named "pdfcompress.sh" and run it with > "sh pdfcompress myoriginal.pdf myconverted.pdf", or make it executable and > put it in your PATH. > > HTH, > /Andreas Hi Andreas
Hmm,, When I try this script, it actually creates a much larger, nasty looking PDF, without selectable text. Also, the script wouldn't work with gsc, but did with just gs. I'm using Ubuntu Hardy Heron, with latest GS, etc. The original PDFCompress script that is on the wiki http://wiki.scribus.net/index.php/Web_optimised_PDF works excellently still for me, and reconstitutes the PDF into an indexable Text too. It really shrinks the PDF from say about 1MB to about 200kb, usually that level of compression, without messing about with reducing image resolution either. It uses much the same techniques, so I'm confused why there is so much difference in functionality. -- Kind regards Julian
