Adobe has announced PDF Print Engine 6, the latest version of its widely used RIP engine, with a number of new and improved features designed mainly to target packaging and textile printing and to take advantage of the larger gamuts that people are using on these presses. The new PDF Print Engine 6 is of special relevance to spot color and extended color gamut printing, and available for software implementation customers in June. Nessan Cleary reports from Fespa 2022 in Berlin.
It’s sold as a Software Development Kit and as such forms the basis of many prepress workflows and Digital Front Ends, with Adobe estimating that it’s used in 80 percent of commercial printing worldwide. The first version, which was first launched at Ipex 2006, was optimized for computer to plate production but since then Adobe has added more features for digital printing and variable data as well as for specific areas such as large format. Mark Lewiecki, senior product manager for APPE, explains, “The big difference being that for computer to plate printing the resolution is much higher but the bit depth is essentially 1-bit screened output whereas for digital printing the resolution is lower but the bit depth is higher.”
The Mercury engine was introduced for version 3, which allows for scalability. He notes, “It’s enabled our partners to build these multi-processor systems. We have partnered with as many as 64 core processors running 64 instances of the PDF Print Engine in parallel. The key attribute of this architecture is that it’s dynamic, meaning that no instance of the PDF Print Engine sits idle while there’s a queue for another one so you cannot predict in advance how long it will take to RIP any given page just because of the mathematics of vector to raster conversion.”
For this latest version, Adobe has made a number of improvements across the board, starting with improved support for extended color gamut inksets. Lewiecki argues that digital printing has led to a growth in extended color gamut printing, partly because more flexo printers are using it to reduce makeready time to compete against digital, but also because it’s very rare for digital printers to offer spot colors and emulating spot colors is much easier with an ECG inkset.
The early versions of APPE assumed that the printing would come down to four plates – CMYK – and relied on a second color management stage to make additional color plates. The PDF Print engine has been able to handle N-colors, including orange, green, and violet alongside CMYK, in a single stage since version 5. But now Adobe has added a number of improvements including transparency blending for ECG inksets.
Lewiecki adds, “Spot color mixing is another area where we have a real strength in that we use CIElab as the profile connection space for spot color mixing and we are also adding the ability to pass through designated plates such as the black plate to basically bypass color management for designated plates so if you have a 50 percent black it won’t get converted to cyan, orange and whatever. The benefits of doing it in a single stage is that there will be greater precision in the mathematics because there are always round off errors in any of these calculations and in one stage there’s only one opportunity for a round off error, so this results in more accurate colors. And the ability to take greater advantage of the full available gamut, which will result in more vivid colors.”
There’s a new spectral module that can process spectrally-defined spot colors alongside the process colors. Lewiecki explains, “We are adding a new API that takes as input CxF spectrally-defined spot colors and converts them so they can be used by the color management system of the RIP. The idea here is that the spot color spectral values and the illuminant, whether its D60 that’s used in commercial print or D65 that is more commonly used in packaging and textile, these are fed as inputs to this API and it produces an output that can be consumed by a color management module or CMM.”
The PDF print engine has its own color engine built in, which is the same as other Adobe programs, such as Illustrator. Lewiecki notes, “A significant minority of our partners do have their own CMM but in either case, these new values plus the input and output profiles are fed into the CMM and that gives you consistently color-managed spectral spot colors just like every other color in the job. So that makes it easier to handle these colors. This API can be used at runtime during RIP’ing or at an earlier stage and some of our partners already have tools like this.”
There’s also a new algorithm that addresses issues with resampling images, typically caused by end customers taking images from a variety of sources, including low-resolution pictures downloaded from websites, which often lack the resolution necessary for commercial print. Previously the PDF Print Engine has got around this with a form of interpolation that just duplicated adjacent pixels.
Lewiecki explains, “Photoshop includes some algorithms for both upsampling and downsampling images that are considerably more sophisticated and we have taken those algorithms from Photoshop and incorporated them into the PDF Print Engine. You could think of this as a form of averaging values of adjacent pixels on all the different color planes, which will result in better reproduction of photographs that are scaled up or downsized, not more detail, but more accurate colors in those interpolated pixels.” He adds, “This is not an AI-based machine learning algorithm although Photoshop does also have a tool like that.”
Adobe has also looked at how presses work with finishing equipment. This includes managing and automating post-press and inline embellishments such as varnish and foils, along with other non-graphic elements, and converting dielines. Typically this would mean a CAD file containing all those converting elements being used alongside the graphics file.
The Ghent WorkGroup has also looked at this, leading to the ISO 19593 standard for PDF processing steps. This specified precisely how to define manufacturing elements such as die lines but also post press elements such as application of spot varnishes and metallics or white underprints. However, this is a very specialized approach and though you can get tools to take care of this, such as Esko Art Pro+ and Axaio’s MadeToPrint Illustrator plug-in, many people just get around this by setting these effects up on a spot color channel.
Lewiecki says that although this approach does work there is room for errors if it’s not properly set up and that could lead, for example, to the effects channel knocking out one of the colors from the graphic. “So we are making these types of elements a first class citizen in the RIP workflow. The way it works is that the prepress operator will map the spot color in the PDF file to a reserved name, giving a formal, official status for the spot color. So it’s no longer a spot color called ‘Varnish’ but is a name that the RIP recognizes as a varnish layer so that you are adding intelligence to the non-graphic elements and enabling the RIP to know in advance how to treat them.”
Adobe has set up three of these reserved names so far, including White Underprint, Varnish, and Die lines. This approach makes it easier to set up different options for the various stages of a workflow so that for example, you can suppress the die lines on press or put them in different colors when previewing the file, or you might want to pass the die lines through to cutting equipment for an offline process. Crucially, this new approach means that all the graphics and non-graphic elements can stay together as part of the same file so that if there are any changes made to any part of the job then those changes are reflected through the whole job, greatly reducing the scope for errors further down the line.
Separately from this, it’s also now possible to automatically generate additional channels such as varnishes, applied on top of other inks to alter their finish, or white underprints, which are commonly used to make a colored graphic stand out from the substrate. That might include DtG graphics printed directly to a dark textile, or labels or window graphics printed onto clear media.
Lewiecki explains, “We know that many of our partners have methods to generate the white plate but doing it inside the RIP is more robust and more accurate with less chance of error and so I’m very pleased that almost all of our partners whom we showed this too, who already have this feature, they agreed that doing it inside the RIP is more robust and is the right way to do it.”
He continues, ”We can choke back the perimeter of the white underprint so that if there’s any misregistration issues then you are not going to get a white halo effect. Also, white ink is the thickest and the most expensive of the inks and there will be an opportunity to dial back the level. Initially, that will be uniform but in the future, we might say that if there’s more ink coverage at a given spot with the design then let’s reduce the amount of white underprint or increase it if there’s less coverage.
“We could mimic the varnish as a behavior for previewing and proofing because designers and prepress operators are basically working blind. They assign a spot color for a varnish but they can’t picture it exactly. The other options might be to show the varnish as an outline or as a highlight color. And at press time if the varnish is not happening inline then you could ignore it but you might also want to pass through the contours to be filled with varnish to a near-line finishing device.”
Adobe has also looked at how to expand support for variable data printing. Lewecki argues that there is a need for short run data – which he refers to as Lot printing, and which covers things such as the ‘Best before’ date – things that don’t change with every item in the way that personalized variable data would, but that do have to be changed with some frequency so that they can’t be included on the plate along with the main design. He adds, “As the runs become shorter and shorter you need to be able to very dynamically and easily put this manufacturing information in, in a very lightweight fashion without going all the way back to Illustrator to change the design, because you are not really changing the design.”
As part of this, APPE 6 will support PDF VT3. This is based on PDF/X-6, which is particularly relevant to multi-page VDP jobs. Lewiecki notes, “One of the key new features in PDF/X-6 that I think is relevant here is ‘Page level output intent profiles’ where the color profile can be different for the cover page from the inside pages.” This is a common requirement given that the cover will often be printed on a thicker substrate, and possibly even a different press. This level of support is included in PDF 2.0 and X-6.
He adds, “The other thing is that for common elements that need to be cached, you can specify that all of their graphic attributes can be encapsulated so that you don’t inherit the attributes that are currently standing in the graphic’s state.”
He suggests a new class of variable data that he refers to as variable product printing, which includes different types of variable data, such as short-run data, the Lot information, different versions with different graphics or languages, product accountability such as serial numbers and bar codes and variable design such as Xeikon VariOne and HP Mosaic. He argues that this is mainly applicable to industrial printing, pointing out, “Variable data printing is a core strength of the PDF print engine. We have this Mercury scaleability architecture, we have parallel processing for the interpretation stage, for the color management stage, for the rasterization stage, and even for the encryption stage and we have automated cacheing including sharing cacheing across multiple instances. So that’s a core strength.”
Fine Line rendering
Adobe has also taken the opportunity to remind everyone that back in 2020 it introduced APPE v5.5 which ushered in a useful new feature – fine line rendering – which was not widely covered at the time. Lewiecki explains that in some packaging workflows the printers require that text be converted to outlines, “So that’s a feature of Illustrator that’s quite easy to do but in changing the text from text, which references a font that has little outlines that describe the glyphs in the font, so when you convert those fonts to graphics it changes how they are rendered.”
This is because Adobe uses a particular technology called CoolType to render its fonts but for graphics, it uses a different rendering algorithm with a different emphasis that’s not as precise. To get around this, Adobe added a Fine Line rendering option to v5.5. Lewecki says that some OEM partners have taken this option. Agfa, for example, makes use of this for its anti-counterfeiting security technology, where fine lines that are difficult to reproduce are commonly used to thwart counterfeiters.
In summary, it’s hard to overestimate just how important the PDF Print Engine is to the printing industry, given that so many RIPs and front ends are based on it. Inevitably we will see a wave of upgrades to various workflows in all sectors, from commercial and packaging print through to wide format and textiles as the various OEMs add this latest version into their products. Some of Adobe’s OEMs already have their own approaches to solving some of the issues we’ve covered here. However, it’s likely that most of those will abandon their own solutions since it’s much simpler to use the technology that Adobe is offering, which by default will also be compatible with whatever Adobe does in future versions of APPE.
For me, I think the real value of this release is that Adobe appears to be once again interested in engaging with the graphic arts and commercial printer community. After all, the foundation behind the entire company was its PostScript page description language. The founders, John Warnock and Charles Geshke, initially came up with the concept whilst working at Xerox’s Palo Alto Research Center, or PARC, and left to set up Adobe to develop it fully. A few years ago I phoned Adobe’s UK press office to ask if they were planning to do anything to mark the 30th anniversary of PostScript but the person I spoke to had no idea what I was talking about and carefully explained to me that PostScript wasn’t an Adobe product. So it is good to see that Adobe is once again interested in professional printing.
Adobe will ship APPE 6 to its software customers starting in June. That would mean that we could see updates to workflows later this year as those vendors incorporate the new version into their software. You can find further details from adobe.com – use this link (https://www.adobe.com/products/pdfprintengine.html) as it’s hard to find this from Adobe’s front page.