Barcodes as machine-readable symbols have been around for years now. They are the smartest way from print to the digital world, but – other than for “normal” print content – not the human eye but a barcode scanner is the one who should be able to read it. That also means that humans are - without tools - not suited to assess a printed barcode’s quality, yet they have to manage their creation in prepress and print. pdfToolbox 12 introduces new technology to determine barcode parameters in PDF files (prepress) and a recent Fogra research project took a close look at quality of barcodes in print. Read Andi Kraushaar’s summary about the Fogra project “Improving the printability and legibility of 1D and 2D codes in inkjet printing” and a few words on brand new pdfToolbox 12’s features for barcodes.
A recently finished research project at Fogra dealt with the improvement of 1D and 2D codes (symbols), which I call here for the sake of simplicity “barcode”. Printing barcodes with HSI (High Speed Inkjet) presses in PDF-based graphic art ecosystems is challenging for print service providers. Practically the service provider can only either fix prepress (data preparation) or press issues (e.g. speed, material selection, orientation to some extent). In the project, we found that music plays in data encoding and below I will provide a short and concise summary. In addition to the project findings themselves several tools were developed, all of which are available at the project page, see link at the bottom of this article.
In typical point of service (POS) applications, barcodes are part of the print content such as product or information carrying displays or 3D displays. Here barcodes come in in all types of quality such as EAN-codes using JPG raster (pixel based) files. In many cases the print service provider is responsible and (financially) accountable for the successful readability of the codes at the final point of use. We know cases where print shop staff is carrying the first print to a point of sale shop just to test its readability. Printing is only continued after confirmation that scanning at the desk was ok.
Often the quality of a barcode is determined based on ISO standards and physical devices, so called “verifiers”, that allow for measuring (not scan) the quality of 1D codes (ISO 15416) or the 2D codes (ISO 15415). The result is an overall symbol grade (OSG). This grade ranges from 0 (fail) to 4 using 0.1 steps.Tab. 1: Numerical values and practical meanings.
|Overall Symbol Grade (OSG)||Meaning|
|3,5 - 4,0|
|2,5 - 3,49||Good|
|1,5 - 2,49||OK|
|0,5 - 1,49||Sufficient|
|Smaller than 0,5||Failed|
In the project, a PDF test form "Fogra Barcode Testform (Eichsfeld) was developed with support from callas which allows for testing all-important parameters influencing the readability of 1D and 2D codes (callas pdfChip was the perfect tool to create the form.) The printed result of the form provides important information already by a visual evaluation.
All technical parameters can be determined by an instrumental evaluation using a verifier. For this purpose, an Excel based evaluation program was created that enables print service providers to determine the optimal parameters for a given printing condition. A video tutorial 'Barcode Property Finder' for this spreadsheet can be found here.
The test form was printed on more than 50 different HSI print combinations and measured with five industry-standard verifiers available. The resulting collection of samples was the starting point for a comprehensive data evaluation.
It turned out that the codes created and printed in true black (CMYK=0,0,0,100) achieved the best results. Of the factors that can be influenced by the print service provider, the bar width reduction (BWR) turned out to be the most decisive quality criterion. Here, an optimum of 30% was identified for 1D codes and 12% for 2D codes. Using the barcode features in callas pdfToolbox the barcodes can be checked to measure input values.
That means that print service providers have only a few options to improve barcode readability: They can check the color encoding and make sure that black barcodes are using CMYK=0,0,0,100 only and secondly, they can check and adjust bar width reduction. If the barcode is encoded as an image, image resolution should be analysed. In case of lossy image compression (JPEG) this has to be combined with a visual inspection. In case of severe problems with such images, it might be necessary to replace the barcode with a new vector based code. This would usually be straightforward in a tool like pdfToolbox, but problematic with colorful barcode backgrounds (which may already pose a problem by themself).
A more sophisticated option is to print the Fogra Barcode Testform using the relevant printing condition and determine the optimal parameters for color encoding, orientation, bar width reduction and module width/size using a verifier. This is explained in detail in the project report.
An additional finding was that even during the course of the project, we saw less and less problems related with barcode readability. We attribute this to a strong improvement of barcode readers that are becoming more and more "merciful". However, that is only partly helpful for print service providers, because their contracts are often based on OSGs instead of problems with real barcode readers, so that an improvement in such readers does not change much for them.
A few more words from callas on this:
In pdfToolbox 12, the interactive barcode reader is part of the “Free Tools” that are available without a license or activation.
The current pdfToolbox 12 release contains technology to recognize barcodes via pattern recognition and allows for creating overview reports for all barcodes in a PDF file with their properties like symbology (type), coordinates, height and width, module width and bar width reduction. Checks allow for finding out whether specified values, e.g. for bar width reduction are kept. It would even be possible to use information encoded in a barcode on a PDF page to adjust the process for the file.