VIGC PDF RIP Audit
Update March 2013: VIGC stops RIP audits (Passed/Perfect labels): ambiguities in PDF specifications
The VIGC has stopped the Passed/Perfect labels for PDF RIPs. The reason: some properties of PDF can be interpreted in more than one way. And there is no unambiguous interpretation to be expected within a short time frame. The VIGC advises all users to keep this situation in mind and to choose for a safe approach, to perform enough tests with their own systems to get consistent results.
When the VIGC started with the RIP audits, PDF and PDF/X-4 seemed to be clearly defined. The current PDF specifications, PDF 1.7 which later became ISO 32000, were published in October 2006. PDF/X-4 was published in 2008, with an update in 2010. It is only last year that it became clear that multiple interpretations are possible for some very specific properties of PDF. And as long as these multiple interpretations are possible, and they are all in conjunction with the current PDF specifications, we have a problem. To be specific: certain combinations of RGB and transparency blend spaces in a PDF can be converted to one color by one RIP and to a slightly different color by another RIP and none can be considered ‘wrong’ by the current PDF specifications. They are just ‘different’.
One of the test files we used for our RIP audits, is the Altona Test Suite version 2 (ATS2). Which is a very thorough and good test file, with a large number of test patches. Next to every test patch there is a reference image which shows how that patch should be rendered according to ECI, the organisation who made the ATS2. Most of these patches are rendered correctly by all RIPs we have tested, but some of them, mainly those that handle transparency, showed differences between RIPs. After many contacts and discussions it was concluded that those patches can be interpreted in multiple ways: there are multiple ‘correct’ renderings possible. Which means that the interpretation by ECI, the reference in the ATS2, is not the only correct interpretation. There are also other, slightly different interpretations possible. ECI explained this recently on their website. (VIGC even saw significantly different interpretation of those patches during the RIP tests, however VIGC did not check whether those significantly different results would comply with the PDF specifications.) Since there is not one unique and correct interpretation, but multiple, the VIGC has decided to stop their RIP audits.
The tests the VIGC has done in the past are certainly not a waste of time. One of the important conclusions is the importance of the correct settings in the RIP. These are essential to get the right output and without the necessary knowledge those settings may cause errors, no matter the capabilities of the RIP itself.
No chage in the short term
Both PDF and PDF/X-4 are ISO standards. Which means that a lot of steps have to be taken before changes can be applied. So it might take some time before the ambiguity in the PDF specifications will be solved. And only then software developers will be able to start adjusting their RIPs and only after that the problem can be solved in the printing industry.
The process of ISO-standards does take some time, because of its thoroughness. But the VIGC hopes that there will soon be an agreement, even an informal one, which can later be implemented into the ISO standard.
The safe way
There is however a ‘safe way’, by avoiding the ambiguous elements. The ambiguous situations are known: combinations of certain transparency blend spaces and RGB. By avoiding these kinds of constructs as a designer, and alternatively work in CMYK when it concerns print jobs, these kind of elements will not end up in a PDF file. That is also the way the GWG specifications will take in the future: the advantages of PDF/X-4 (transparency and layers), but no RGB.
------ end update March 2013 -------
The next big step in prepress will be PDF/X-4, which allows transparency and layers. But what about the support of PDF/X-4 by RIP vendors? VIGC has carried out its first 'PDF RIP audit' and is pleased to report that seven systems were awarded its 'PDF RIP Audit: Passed' label. The audit involved the first large PDF/X-4 RIP test in the world, and following the results a number of vendors have improved their products.
Test set: Altona, new GWG patches, VIGC color test files
The test set we used wasn’t made up of average documents. On the one hand, we used Version 2 of the Altona Test Suite, which has a huge amount of possible combinations, some even very uncommon. On the other hand, a number of new patches from the Ghent Output Suite (GWG) were used – a large number of those were developed by VIGC. And to test the possibilities of color conversions, we also created additional test files.
The problem areas – transparency blending spaces and black point compensation
The main conclusion of the VIGC PDF RIP Audit highlights the complexity of transparency. During the tests, there seemed to be differences in interpretation of some very specific cases, which can be traced back to different interpretations of that small part of the PDF specification itself.
PDF RIP Audit – Passed and Perfect
We award a label to the systems that have passed our tests. We make a distinction between a ‘perfect’ accreditation – where everything was rendered as expected – and 'passed'. The difference between the two is mainly a part of the Altona Test Suite v2, i.e. that part where there’s disagreement about the correct rendering of ICC based colors in combination with transparency blend color spaces. We have taken the ATS v2, untill further notice, as the reference. Also, some patches of the VIGC test files weren’t considered 'wrong' – for example, when minimal color differences occurred.
In both cases, Passed and Perfect, the systems awarded with these labels are 'fit for producton'. The minor differences between both labels cover very specific test elements, which are rather uncommon in real life.
The following systems received the VIGC PDF RIP Audit: Passed label and are considered 'fit for production' (30/04/2012):
- EFI - XF v4.5.5
- Esko - Esko Suite 12
- FujiFilm - XMF v126.96.36.199
- GMG - ColorProof v188.8.131.52
- Heidelberg - Prinect 2011 SR1
- Kodak - Prinergy 184.108.40.206
- Xeikon - X800 Digital Frontend
Update 8/10/2012: just before Graph Expo a first system got a perfect score, which means that it also rendered all the test elements of the Altona Test Suite v2 as expected by the makers of the ATS2.
- EFI Fiery, versie FS100 Pro
The following systems got a 'passed' label just before Graph Expo:
- Agfa :Apogee Prepress v8 (AP 8.0.0b3)
- Océ PRISMAsync Color Controller Versie 13
We have drafted an extensive test report featuring the results of those systems that passed the audit, as well as the issues encountered with other systems (note: these systems haven’t been named.). The full report can be purchased from the VIGC, the cost is 250 euro (excl. VAT and shipping, payments via credit card or paypal).