I wondered if anybody could explain what differences there are in the PDF which would differentiate

  1. pages produced by running pdflatex on the .tex file where the code is part of that file


  2. pages produced by running pdflatex on the .tex file where the pages are the result of \includepdf and the included pdfs were produced by a separate run of pdflatex on separate .tex files.

For example:

Here is a page of the first kind.

Where pdfpage.pdf is the result of running pdflatex on

Here is a page of the second kind.

I am interested because although my compiled documents print fine on printers, if I try to print from a photocopier, it will not print anything after the first includepdf. (If there is nothing before this, it gives an error; otherwise, it just prints the pages which came before that point.) For various reasons, I have to print some things from the photocopier rather than a printer and so this is obviously quite awkward. The problem is not specific to any particular document but appears to be general to documents I produce in this way.

So what might the photocopier but not the printer see or not see at the point where \includepdf occurred in the source? (And is there anything I can do about it?)

Note that the MWE is not a real one because that would require me to run test documents on machines at work and I would prefer to do that only if I cannot learn anything about the problem in any other way. So the example represents the pattern I've noticed but I cannot rule out something else being the real issue at this point.

  • 4
    For a quick fix for the problem with the photocopier, you might try running the PDF through Ghostscript (if you don’t want to learn the command-line syntax, just use pdf2ps followed by ps2pdf). This usually fixes all types of troublesome PDF files. (Note that the problem isn’t necessarily with the actual PDF file, but may with the photocopier or its drivers, but this solution will likely still work.) Jan 26, 2014 at 9:54
  • Thanks for that suggestion. I'll give that a go. For some reason, that hadn't occurred to me even though I've used the pdf2ps -> ps2pdf trick in other cases. (Or even just pdf2ps although that wouldn't work with the photocopier.)
    – cfr
    Jan 26, 2014 at 14:09

1 Answer 1


For every page of included pdfs pdf/Xe/LuaTeX creates a "Form XObject" (PDF Reference, section 8.10), which introduces one more "level of indirection" than for directly created content (which resides "directly" in the Page Objects (section This additional level of indirection can cause problems in older Raster Image Processors, as it is one additional "q/Q nesting" (see section C.2).

  • Thanks. That's helpful. (I'd at least like to understand why I can't do what I want!) What's RIP? This is a brand new photocopier, though I doubt that's worth much. Is there any way of post-processing such PDFs to avoid the problem? I tried the pdf2ps -> ps2pdf trick suggested in the comments and it helps but doesn't solve the problem. (Now some included PDFs get printed but not, alas, all of them and I have no idea why or which...) Is any particular specification of \pdfminorversion= better or worse in this case?
    – cfr
    Jan 28, 2014 at 2:30
  • @cfr: RIP = en.wikipedia.org/wiki/Raster_image_processor. There's not much you can do but file a bug report with the manufacturer. Good luck there. :-{ Jan 28, 2014 at 8:19
  • Thanks. (RIP means a lot of things so thanks for disambiguating.) Hmm. On the one hand, at least I know it isn't my fault. On the other, it sounds as if I'm stuck with the problem. At least I have some idea why now. That may sound unhelpful but I generally deal better with things I understand at least a bit even if I can't do much about them! Thanks, too, for the PDF Reference link although I think a lot of it will go over my head...
    – cfr
    Jan 28, 2014 at 17:39

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.