I often use Ghostscript (in Automator) to try to pre-process vector pdfs (so they're faster on my Mac, and compatible with my Kindle). Even just running -sDEVICE=pdfwrite without the rest of the pre-processing often creates gray or black rectangles, which sometimes obscure text. For a mild example, it adds a gray box to the title of the rules here: https://www.peginc.com/store/genre-supplement-wizards-warriors/ Adding -dFILTERIMAGE does *not* remove the rectangles, and sometimes removes good text.
Adding -dFILTERVECTOR *does* remove the rectangles. But it is just a kludge.
(In reply to MarjaE from comment #0) > I often use Ghostscript (in Automator) to try to pre-process vector pdfs (so > they're faster on my Mac, and compatible with my Kindle). > > Even just running -sDEVICE=pdfwrite without the rest of the pre-processing > often creates gray or black rectangles, which sometimes obscure text. > > For a mild example, it adds a gray box to the title of the rules here: > > https://www.peginc.com/store/genre-supplement-wizards-warriors/ Please attach example files here. URLs often go stale before anyone has time to look into the problem.
The file in the URL is a Zip archive, which contains 3 more zip archives. After extracting all of these I end up with 7 PDF files, its not clear to me which of these are the 'rules' you refer to. One of them isn't even a PDF file.... span.pdf contains HTML. Pretty much all of these 'vector' PDF files contain images, several of them contain JPEG2000 images. In the absence of a full command line, I've run all of them through Ghostscript with a simple command line: gs -sDEVICE=pdfwrite -o out.pdf I do not see any problems with any of the (numerous) pages in any of the PDF files. Of course, in a large number of pages spread across several files, its easy to miss problems. You need to supply a file and a command line which will reproduce the problem, you also need to say how you are opening the PDF file; have you considered the possibility that the problem is with the PDF viewer you are using ?
Created attachment 15634 [details] One affected file
I run from automator using the following: Run Shell Script Shell: /bin/bash Pass input: as arguments for f in "$@" do suffix="-converted.pdf" base=`basename "$f" .pdf` outputfile=$base$suffix /usr/local/bin/gs -sDEVICE=pdfwrite -sstdout=%sstderr -dNOPAUSE -dQUIET -dBATCH -sOutputFile="$outputfile" "$f" done *** I get the same results running from terminal using the following: gs -dNOPAUSE -dQUIET -dBATCH -sDEVICE=pdfwrite -o test.pdf [input file address] *** I see the same gray boxes in the Finder preview window, in Clearview, and in Preview. In the sample file, I get a gray box around the title Wizards & Warriors on pages 1 and 8.
(In reply to MarjaE from comment #5) > I see the same gray boxes in the Finder preview window, in Clearview, and in > Preview. In the sample file, I get a gray box around the title Wizards & > Warriors on pages 1 and 8. I assume you mean pages 1 and 6, since page 8 doesn't have this on it (whereas page 6 does) I don't see any problem with Adobe Acrobat Pro, Ghostscript, MuPDF, Google Chrome or another PDF rendering engine I have to hand. I have no idea what Clearview is, but the others are part of MacOS and use the Quartz rendering engine. If there's a bug in that, then they will (clearly) show the same result. Quite likely any MacOS application will, because they will use the system libraries to render the PDF and. if those libraries have a problem, then all the applications using those libraries will display the file incorrectly. Try opening the file in Adobe Acrobat Reader, my suspicion is that this is a transparency bug in the rendering engine.
I don't have Adobe Acrobat Reader. I can't download from their website, since I'm photosensitive, and their site is prone to reload and flash every 1-2 seconds.
Created attachment 15636 [details] Same affected file after running through pdfwrite
Kindle Dx cannot show images in original, has 3 gray boxes on each of pages 1 and 6. Can show images in processed version, still has 3 gray boxes.
(In reply to MarjaE from comment #8) > Created attachment 15636 [details] > Same affected file after running through pdfwrite This renders perfectly in every PDF consumer I have. The only Mac I have access to is running 10.8 (Mountain Lion). On this Preview shows a (very faint) rectangle behind the title, and an equally faint rectangle behind the image of the characters. I don't see this on any other PDF consumer. I suspect you have found a bug in Quartz and should report it to Apple.
I get the same gray box on my Kindle, which doesn't use Quartz.
Comment on attachment 15634 [details] One affected file I posted the attachments by mistake; the file was free to download but not free to redistribute. Is there any way to remove it?
(In reply to MarjaE from comment #12) > Comment on attachment 15634 [details] > One affected file > > I posted the attachments by mistake; the file was free to download but not > free to redistribute. Is there any way to remove it? Not really, no. I can (and have) mark the attachment as private which will prevent anyone outside Artifex accessing it, but that's all. I'm dubious of the value of that after 6 years though :-)