Even though PDF is a reliable file format, things can still go wrong when processing PDF files.
- Sometimes a PostScript error pops up and a PDF file cannot be output.
- Acrobat occasionally doesn’t seem capable of displaying the file correctly.
- There can be inconsistencies between the output and what a PDF looks like on-screen.
- Older software may not be capable of handling the features of the latest PDF versions.
A frequently used trick to try and get around such problems is ‘refrying’ the PDF file: converting a PDF to a PostScript file which is then converted to PDF again. This ‘second generation’ PDF file often does not exhibit the problems with the original source file.
Some people print the original PDF to a PostScript file and then distill that file. Others are firm believers that you should export to PostScript from within Adobe Acrobat and then convert that file to a PDF again. Both the Print and Export menu options actually use exactly the same software routines in Acrobat so it does not really matter which method is used.
Other reasons to refry PDF files
Some users refry PDFs because of the way they process data:
- PDF-based advertisements are placed on InDesign or QuarkXPress pages.
- These pages are printed to a PostScript file.
- This file is sent to Acrobat Distiller or another PDF creation tool.
It is not always apparent to users if a PDF file goes through an intermediate PostScript phase: if a PDF is placed on a QuarkXPress, PageMaker or FrameMaker page which gets exported to PDF, that PDF gets refried because these layout applications internally always converts PDF to EPS (Encapsulated PostScript).
Last but not least: some people refry every single PDF they process, either because one of the above reasons forces them to do so or because refrying solved a few problems in the past and now they assume it won’t hurt to do it all of the time.
Is it a good idea to refry PDF?
As a way of troubleshooting files: YES, but check the remainder of this page to avoid creating more problems than get solved.
As a fixed part of your workflow, even though there is no technical reason for doing so: NO! Not only are you wasting system resources but each transformation brings along the risk of data loss or something going wrong.
What can go wrong when refrying PDF files?
PostScript is an older file format than PDF. The differences between both systems cause some transformations when refrying PDF files.
- When it comes to color management, PostScript is CIE-based whereas PDF relies on ICC profiles. As long as you are working with CMYK-based PDF files (which probably account for over 90% of all PDFs out there), there is no need to worry about color shifts or changes when going from PDF to PostScript and back. If a PDF however contains color managed data (e.g. it is a PDF/X-3 file), the conversion to PostScript forces a conversion from ICC to device or CSA based colors. The current Adobe applications use a work-around to preserve the ICC profiling data. They embed the ICC-profiles in DCS comments which can be picked up by Distiller again when converting from PostScript to PDF. Other or older applications are not that clever and colors may change when refrying PDF files.
- Since PostScript doesn’t support transparency, refrying PDF files implies that the second generation PDF won’t contain any transparency. That means this file may be larger and can’t be edited as easily any more. Nowadays both Adobe and Global Graphics ship renderers or rips with full support for transparency. It is a shame not to use that capability and revert to flattened PDF files.
- Both PostScript 3 and PDF 1.3 & later support a mechanism called smooth shading which offers higher-quality monochrome or color gradient fills. Older versions of PostScript as well as the EPS file format don’t support smooth shading. This means that gradients are transformed when converting from PDF to EPS or PostScript level 2. This may lead to banding or thin white lines in the gradient.
- A minor inconvenience for prepress users is the fact that a number of PDF features such as annotations, forms, hyperlinks, movies or bookmarks won’t survive the transition to PostScript.
- There are some lossless compression algorithms that are supported by PDF but not by PostScript. Refrying a PDF file that uses such an algorithm can lead to significantly larger files. This can also happen with PDF files in which JPEG2000 compression is used. This algorithm is more efficient than the regular JPEG compression that is supported by PostScript.
- I have never encountered PDF files with 16-bit images in them besides the ones that I created myself when playing around with that feature. Those 16 bits won’t survive refrying. A refried PDF always contains 8-bit data because that is all that PostScript can handle.
Incorrect settings can affect the refried PDF
- Both the settings used to export a PDF to PostScript and to distill or normalize that PostScript file back to PDF affect the quality of the refried PDF. If Distiller is set to downsample images or recompress using a low quality JPEG setting, this can obviously affect the quality of images.When exporting or printing to a PostScript file, make sure to use the following settings:
- PostScript: select “Language Level 3″.
- Fonts: deactivate any option that converts fonts & make sure all fonts are embedded.
- Color management: deactive it (“Same as Source” option in Acrobat 8).
- Transparency: set the Flattener Preset to “High Resolution”.
In Distiller or any other PDF tool that is used to create the second generation PDF file, use the following settings:
- Images: deactivate downsampling & don’t try to compress the images too much. Activating the “Save original JPEG images in PDF if possible” option in the Advanced settings of Distiller 8 is a good idea.
- Color management: deactive it (“Leave color unchanged” option in Distiller 8).
- Due to incorrect settings, fonts may get converted to another font format or they may get renamed and subsetted. There is always the risk of losing part of the editability of the file because the Acrobat Touch-up tool doesn’t work that well with a refried PDF.
- Every time data are transformed, there is a risk of losing or damaging something. Some font related issues which may pop up when refrying PDFs or merging multiple PDFs in a single file are related to software bugs. Characters that vanish, get replaced by others or appear as small square rectangles are typical examples of bugs in applications that do not properly handle the merging of font subsets. This is actually the most fearsome problem in any workflow in which PDF files are reprocessed. Imagine the first digit of prices in a catalog being replaced by spaces! I once heard of a printer having to reprint an entire series of math books because a few obscure characters with an important mathematical significance somehow got replaced by other funny looking characters. No prepress operator can catch such an error (unless you happen to be a math buff with time to spare to check hundreds of page proofs).
Alternatives to refrying
Try correcting the original PDF file instead of using the more drastic refrying procedure. This is both faster and the risk of losing or altering data is lower.
- From Acrobat 6 onwards, Acrobat Professional has a ‘PDF Optimizer’ function which can do a good job of cleaning up a file. In Acrobat Professional 8, this option can be found in the ‘Advanced’ menu option.
- Third party preflight tools or optimisation tools may also be able to solve the problem. Enfocus PitStop, Callas pdfCorrect & Apago PDF Enhancer are examples of such tools.
This page is my interpretation of a discussion organised by the Ghent Workgroup. Leonard Rosenthol and Dov Isaacs from Adobe provided a lot of the above information. Olaf Drummer (Callas), Jo Brunenberg, Stephan Jaeggi (www.prepress.ch) and other GWG members provided additional information and insights. Meanwhile GWG have published a document about refrying PDF with some interesting recommendations.