by Peter Kleinheider,
Manager Print Publishing Technologies,callas software GmbH

Since many years, the PDF format reached its final breakthrough in the transmission of finished print data. Therefore, the question whether open files or even PDF are to be used is no longer valid. The variety of problems that occurred earlier in connection with native software layout formats were drastically reduced by PDF. However, the PDF is not the panacea for all problems with print data, which we all have experienced first hand. The PDF format by itself is dumb and does not contribute towards a high quality of the file content to be printed. Rather, it can be considered as a container where it depends on how it is filled.

Some of the negative examples are non-embedded fonts, colors not corresponding to the requirements or wrongly defined format dimensions. As the creation process of a layout cannot be checked in this step, at this point, if you are working with the most varied source material and person groups, any errors can be counteracted by an appropriate test of the printable PDF.

To stem the wild growth of PDF test profiles and to give the industry a solid basis for basic requirements for printable PDFs, the Technical Committee TC 130/WG 3 of ISO composed. In the year 2002 (among other things, due to strong support from Germany and Switzerland), it baptized the Standard PDF/X (ISO 15930). The X stands for “blind exchange” – i.e. “without necessary queries”.

In the meantime, it was 2007. The PDF/X Standard was once again adapted, PDF files were almost exclusively used for delivery of printable data, in fact, the ratio of PDF/X files is abysmally low. Did the problem-free PDF processing not achieve the set goal?

When you overhear “only” PDF suppliers, there are many excuses to avoid delivering PDF/X files. So:

• What is PDF/X?
• We would love to supply it, but our printer does not support it?
• We tried it, but there were problems?
• When we want to create a PDF/X, we constantly get error messages?
• For PDF/X, we need an ICC profile, we are not yet there?
• PDF/X requires an ICC profile for definition of the output intent, however, the printer does not give us any profile?

One is right in asking: What went wrong with PDF/X? For me, the questions indicate that there is a lot of misunderstanding among users.

There is a very simple answer to the question “What is PDF/X?”: “A PDF/X is a simple PDF file that was checked right up to whether only printable elements are included, whether all necessary elements are available for output and whether all colors were defined. Moreover, a PDF/X file transports the intended output condition.” Hence, a PDF/X file must be processed in every PDF-capable workflow without further problems, as it “merely” deals with a tested and PDF file marked according to the PDF/X guidelines.

As for the problem of the ICC profile necessary for the PDF/X creation: The so-called OutputIntent defines the print procedure for which the data was prepared. As a rule (or in case of doubt), it pertains to data that is prepared for the offset printing on coated paper. The OutputIntent to select is hence ISOcoated V2 - the standard profile for this print condition. This ICC profile does not change the data under any circumstances and can be used without doubt. To enable the recipient to rely on this information, the correct ISO profile must be selected for other print conditions (i.e. for printing processes other than offset printing on coated paper).

For the statement “We tried it, but there were problems?” I must ascertain: PDF/X hardly makes a statement about the resulting quality. Further-reaching requirements such as total ink application, name and number of spot colors, minimum picture resolution, minimum rule weight, page dimension, number of pages and trappings are not defined by the standard. As all these parameters greatly depend on the print procedure, the desired quality and the printing substrate, defining these requirements in a standard makes no sense. These definitions would be too multiplex and too fast moving, the standard would not be applicable. For these purposes, Industry Associations have defined specifications and quality guidelines for various output intents. In this context, the international Ghent PDF Workgroup (short GWG) and the Swiss PDFX-ready Initiative (based on the GWG specification) are worthy of mention. Both have the task of defining additional criteria based on the PDF/X standard, which presuppose the qualitative and typographic interests of different output processes.

The characteristic of PDF/X as PDF/X-1a (ISO 15930-1 and 4) and PDF/X-3 (ISO 15930-3 and 6) creates some confusion. Even here, the answer is simple: If only process and spot colors were used during the design stage, the PDF/X-1a standard hits the spot. However, if you want to define your media independent RGB data with specified source profile even in the PDF, then the PDF/X-3 standard is the right thing for you. However, we must add: A PDF/X-3 file must not necessarily contain defined colors per ICC profile. There is only one option. In Germany, Austria and Switzerland, the methodology of the general PDF/X-3 delivery was established, however with the restriction that no ICC based colors are to be used.

PDF/X files are either generated via interactive PDF verification tools (e.g. the Preflight tool integrated in Acrobat Professional), during direct export from the layout program (e.g. InDesign PDF export), or as part of the PDF creation via Acrobat Distiller. In automated PDF workflows, PDF/X generation and verification modules are likewise used if these are configured accordingly.

To proceed safely before the final output, the PDF must be printed once again (with Simulate Overprinting enabled please), and a separation check must be conducted in case of spot colors and overprinting elements. Now, if you have a PDF/X file that in addition corresponds to an industry recommendation, then problems during output are largely ruled out.

On the recipient side, this guarantees more effective data processing, as some warnings were already sensed by the suppliers when testing the PDF, which minimizes queries and consequently saves time too. If the PDF is printed 1:1 in the next course, a 100% match with the visually checked result is guaranteed on the supplier side.

The PDF/X-4 standard has been defined due to the increased use of transparencies and the desire to preserve these in the PDF. The PDF/X-4 standard is based on PDF/X3 and in addition to transparencies, enables JPEG2000 compression and use of layers. It might be published in the third quarter of this year. Subsequently, transparencies must no longer be considered flat during the PDF creation, which leads to qualitative losses in case of wrong reduction settings. Even the subsequent conversion of spot colors to CMYK which involved a visual change in the appearance was not always possible. If no modern workflow with automatic transparency calculation is available for the output, I consider the preservation of transparencies practical, as with Acrobat 8, we have a professional tool to flatten transparencies in a controlled manner. Moreover, when working with media-neutral contents in the PDF, i.e. if images in the PDF are also provided with ICC source profiles, then no problem-free output can live without PDF/X-4.

Therefore: If you like to use PDF, please use a tested PDF/X or even better, use a PDF/X in compliance with the guidelines of PDFX-ready or GWG.