I generate my pdfs using MikTex 2.9 on Windows 7. On W7 everything looks good. On my other computer where I have CentOS release 5.2 (Final) installed I experience strange display on page numbers. On that computer I use Adobe Reader ver 9.4.2 02/11/2011. Images below are examples of how it looks like on Page Navigation toolbar:
(this should be page A)
(this should be page 3)

In my document I include 2-page pdf at a beginning (pages A and B). Next, there are tex-generated pages (pages 1 to 17). My input tex document is saved using Unicode encoding.

I know this might not be latex issue, but can I do something in it, so I don't get those glitches on CentOS version of Adobe Reader?

  • did you run xelatex?
    – user2478
    Nov 21 '11 at 10:28
  • 6
    For me it looks as if the page number are in unicode and the reader can't extract them correctly. If you are using hyperref you could try \usepackage[pdfencoding=auto]{hyperref}. Nov 21 '11 at 10:42
  • @UlrikeFischer, your solution worked. You may add an answer so I can accept it and vote up :).
    – Ichibann
    Nov 21 '11 at 19:48
  • @Herbert, I'm using pdflatex.
    – Ichibann
    Nov 22 '11 at 7:30
  • @Ichibann: then follow Ulrikes advice ...
    – user2478
    Nov 22 '11 at 8:06

For me it looks as if the page numbers are in unicode and the reader can't extract them correctly. If you are using hyperref you could try \usepackage[pdfencoding=auto]{hyperref}.


You’re probably using the hyperref package with the unicode option, e.g.:


Just remove the unicode option, and everything will be fine. Note that you can still use UTF-8 as the input encoding.

  • You are right. I use hyperref with unicode option. I don't want to disable it. Is there other way?
    – Ichibann
    Nov 21 '11 at 11:03
  • @Ichibann What is the reason you don’t want to disable it? Nov 21 '11 at 11:08
  • I want to keep Polish diacritics in pdf's bookmarks.
    – Ichibann
    Nov 21 '11 at 11:13

This could be a font embedding issue. If you have access to the pdffonts tool on CentOS you can use it on your pdf to see which typefaces is uses versus which it has embedded. You can probably use Acrobat Viewer to do this too.

If it's missing some, then you can configure pdflatex (assuming you're using that) to embed them for you. I'm not sure about this on Windows, but on Linux the command is

updmap --edit

and it seems on Windows to be

initexmf --edit-config-file updmap

and then find pdftexDownloadBase14 and set it to true. Make sure to check the file to ensure that it doesn't get turned off again.

The next part is shamelessly pinched from an answer to the Stack Overflow question How do I embed fonts in an existing PDF?. Ghostscript can be used to reprocess the pdf to contain the fonts that are missing. This way you can repair existing pdf files that are causing you trouble.

gs       -sFONTPATH=/path/to/fonts:/another/dir/with/more/fonts       -o output-pdf-with-embedded-fonts.pdf       -sDEVICE=pdfwrite       -dPDFSETTINGS=/prepress        input-pdf-where-some-fonts-are-not-embedded.pdf
  • Please, see my edit.
    – Ichibann
    Nov 21 '11 at 9:19
  • Try my solution, then see if it works. ;)
    – qubyte
    Nov 21 '11 at 9:21
  • I've added a ghostscript command. This should allow you to repair pdf files that you have already generated that have this problem.
    – qubyte
    Nov 21 '11 at 9:37
  • What font path is there on CentOS?
    – Ichibann
    Nov 21 '11 at 9:55
  • chkfontpath --list should tell you where they are, but you may need to install that utility. Honestly, I know little about CentOS.
    – qubyte
    Nov 21 '11 at 10:02

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.