Should astronomy journals accept PDF figures for publication?

PS and EPS figures have been the bread and butter of astronomy for decades. In turn, the astronomy journals accepted the PS and EPS figures for publication. Recently, plotting packages are supporting PDF files more than the EPS and PS formats. An EPS file was originally developed as a language to tell a program how a page should look.

You could look at the PDF format as the next step in the evolutionary chain to the PS/EPS format. PDF files can contain additional information that EPS files cannot such as transparency, better support for character encodings, ability to embed TrueType and OpenType fonts, and more. With the more modern plotting packages, you have more flexibility about your figures when saving in PDF format.

Surprisingly, PDF figures are not accepted by several astronomy journals, but there are exceptions like A&A and MNRAS. Because of this many researchers are investing a lot of time to convert PDF figures to PS/EPS format for publication. Because the conversion between PDF to PS/EPS is not perfect, getting the EPS file to look right can be frustrating and time consuming.

Hence, a few astronomers (David Hogg, Thomas Robitaille, August Muench, and Eli Bressert) got together to start a petition to let the publishers know that many of us prefer to give our figures in PDF vs EPS format. We don’t want the journals to cease support for PS/EPS figures, but at least accept PDF figures as a viable format. If you want to join us in getting this message across to the publishers please visit this link to support the petition.

6 comments… add one
  • August Muench Nov 6, 2012 @ 13:14

    Actually, Oxford (the new publisher for MNRAS) prefers EPS though TIFFs are okay too. In fact they basically hate pdfTeX…

    • Matthijs van der Wiel Jan 24, 2014 @ 13:03

      I am not sure if the MNRAS instructions have changed since late 2012, but now (Jan 2014), that exact same webpage with instructions for authors includes the suggestion to add
      to the preamble of your LaTeX file, to avoid issues with their manuscript management system. I hope this trick will work, because I too prefer pdf figures.

  • Médéric Boquien Nov 6, 2012 @ 17:50

    The problems with EPS and PS files do not stop there, unfortunately. They also tend to have a ridiculously inflated size. This can be a problem when submitting figures for an article when you have to send hundreds of MB and the journal website systematically cuts the connection after a few tens of MB. This happened to me and a colleague twice over the past year.
    Honestly I do not know what is the problem with PDF files. As far as I know it is the standard format in the printing industry, and I am pretty sure that offset printers accept PDF.

  • Wolfgang Kerzendorf Nov 7, 2012 @ 9:37

    I have tried before to specifically ask ApJ on that issue. Here’s the email and their response:
    Dear ApJ-Team,

    Would it be possible to also accept PDF Figures for papers? Most modern plotting programs are optimized for PDF and they are more easily viewed on desktop operating systems. The PDF format also has support for transparency, which EPS apparently lacks.

    I have researched where ADOBE (the author of both formats) stands on this issue:

    Some comments from this article include:
    ‘So PDF is a file format that is smarter than EPS. ‘
    ‘The moral to this story is that PDF can be used as a replacement file format for EPS, and that PDF can be used as a delivery format for sending complete publications to press. It is also suitable for soft-proofing, distribution on the Internet, and file archiving, as it is completely self-contained.’

    Many of my young colleagues also feel that EPS is an outdated format and that it would be more efficient if we could also submit PDF figures.

    Thank you for considering this request,
    Wolfgang Kerzendorf

    Their response (Chris Biemesderfer):

    Wolfgang, thanks for your email, and for addressing this concern on behalf of your colleagues. What we really want to obtain from plotting programs are the vector drawing instructions. As you suggest, as long as the plotting software distills the PDF so that those are captured, the PDF can be in principle as useful as the original PostScript. There are, however, many original formats that can be used to create PDF. It is the *original* format that dictates what our preservation policies towards the object are, and it can be difficult to know what those might have been.

    The article on the Adobe site is written for shops where the pre-press environment is known to be PostScript-based. An argument can be made in that case for PDF in favor of PostScript. As I alluded to above, we don’t always know what the original format is. There are cases where the distillation into PDF obscures its identity, and that creates preservation difficulties in our workflow. I will ask our preservation providers whether there have been any enhancements on the preservation side, but for now, we still prefer EPS for vector graphics.


    There was a last email, where it was alluded to, that AAS requires EPS for preservation purposes.

    Hope that shines some more light on this

  • Laura Watkins Nov 10, 2012 @ 6:41

    As August mentioned above, the problem with MNRAS isn’t PDF figures but pdfTeX. To include PDF figures, you need to compile your paper using pdfTeX, and the MNRAS software doesn’t play nicely with papers compiled that way. (I tried anyway, it didn’t work.)

    I managed to get around this by compiling my paper with pdfTeX, opening it in Preview and then using File > Save As and selecting PDF. This made the file size ~3 times larger, so it obviously changed something behind the scenes. When I tried uploading this new PDF to MNRAS, everything worked just fine. So in the end I submitted my original PDF figures and the re-saved PDF paper. MNRAS haven’t complained so far, but I’m only part-way through the whole procedure, so I’m not safe yet.

Leave a Reply

Your email address will not be published. Required fields are marked *