10 May 2017
By Bart Fret
For a long time, colour management was only seldom used in large-format printing. Sign shop personnel would linearize and then send their cyan, magenta, yellow and key/black (CMYK) jobs to their digital printers. As such, they relied on these devices’ printable gamut, but this led to issues with colour consistency and repeatability.
More recently, colour management has gone mainstream in large-format printing, as it is seen to provide a lot of benefits to both print service providers (PSPs) and their customers. There is more than one way to implement it, however, depending on the approach a shop takes to standardization.
Standardizing colour management
The operation of a large-format inkjet printer depends on a raster image processor (RIP). RIPs are always set up with printer profiles, which include ink limits, linearization and, optionally, an International Color Consortium (ICC) profile. When graphic files are submitted to the RIP, the input colours—whether CMYK or red, green and blue (RGB)—are converted to the printer’s output colours, based on tagged image and file elements of the input profiles.
This is the standard colour management process, but numerous problems are associated with it, as follow.
Printed colours may appear different than intended, particularly if the input profile used within the print shop does not match the profile the original designer used. If an image has text layered on top of it, for example, and the colour of the text is derived from a part of the image, then when the file is processed, any change to the profile or intent of the image will also change the colour relative to the text, in a way the designer likely did not intend.
Applying different profiles and/or render intents to different elements, such as images and vectors, can also significantly affect their printed appearance.
Overprints, transparencies and blends
Applying profiles can dramatically change overprinting elements. When converting the cyan and magenta (C+M) elements from the General Requirements for Applications in Commercial Offset Lithography (GRACoL) profile to the printer profile, for example, the conversion may result in not only C+M, but also a bit of black (K), due to the printer profile. At this point, any underlying black text would disappear, because K was generated in the C+M overprinting elements.
Colour differences can also occur with transparencies, as elements that would blend might end up being converted using different rendering intents. Overprints can seem to disappear and blends can look different when assigning profiles.
Perceptual intent changes all colours in a file, not just those that are out-of-gamut. So, colours might look very different, depending on whether they are converted to a standard such as GRACoL or to the printer profile. The same job printed on multiple machines will look very different, since the gamut of those devices will differ and the perceptual intent adjusts all colours to the printer’s output profile.
Undefined spot colours are printed using an alternate space, usually CMYK. If the designer has taken this into account, then the results might look very different, due to the same reasons cited above for colour differences.
By way of example, Figure 1a (right) shows the output of a document created using Adobe Illustrator graphic design software, comprising an RGB image of an acorn, an embedded standard RGB (sRGB) colour space profile, a grey background sampled from the image, text with colour sampled from the image and another gradient derived from the same colour.
Figure 1b shows what happens when the designer’s intent is ignored—likely unintentionally and relying on default application settings. A colour difference appears, for example, because the background is defined in CMYK without a profile. Adobe Illustrator would have defined the CMYK from the grey of the sRGB space and converted it to the Specifications for Web Offset Publications (SWOP) profile, i.e. the designer’s working space. If the combination of profiles and rendering intents assigned to the image and the background is not exact, then such a colour difference will appear.
Further, the acorn became brighter and more saturated compared to the rest of the design. This is because it was mapped directly to the printer gamut. This is a desirable result, but breaks the match between the text and the gradient colours in the design. The gradient is defined by an overprinting spot colour. The background now has two different appearances, also seen in the overprinting and blending effects.
Finally, Figure 1c shows what happens when the spot colours are replaced as RGB or CMYK before being RIP’d or when the overprint is turned off to counter the gradient effect in Figure 1b. The text no longer overprints, as its colour has been replaced with a CMYK version that represents the spot colour. The CMYK match is very accurate, but due to the overprinting effect, the spot colour is darkened by the grey background.
The gradient no longer shows the correct overprinting effect. The original spot colour was a duotone, fading to zero of another spot colour. As a result, it has nicely blended away. Now, however, due to either turning off the overprint or replacing the spot colours with their CMYK equivalents, the overprinting blend is lost and blends to white. Any change made to the original design by assigning or removing embedded profiles, disabling overprints or replacing spot colours before RIP-based blending will result in ‘broken’ output.
With all of this in mind, one may well wonder, why use colour management at all? The alternative of not using it, however, means sending a design file’s CMYK data directly to the printer via linearization, which results in the least colour consistency, so it is definitely not recommended. Rather, the better approach is to use standardized printing.
The goal of standardized printing is to reduce the colour deviations and variations in graphics produced by different devices. These differences usually exist because machines vary in their colour capabilities—sometimes quite dramatically—due to a variety of printhead technologies, ink formulations and substrates.
Standardized printing also respects the designer’s intentional (or unintentional) overprint, transparency and blend effects by converting colours correctly to a standard before the conversion to the printer’s own colour space. This approach ensures the output is accurate and uses everything the printer’s gamut has to offer. As there is no need to test the printer to see if the job will be handled correctly, this saves time for the operator who is preparing and printing the job.
The way to standardize printing is to convert all incoming colour spaces—whether CMYK or not—to a chosen reference profile first, like GRACoL, and only thereafter to the printer profile. Thus, all of the colours are first matched to a standard, then converted with the printer’s specific capabilities in mind, which helps maintain overall colour appearance between different devices.
A subset of the International Organization for Standardization’s (ISO’s) Portable Document Format (PDF) specification known as PDF/X was developed for this reason, among others. PDF/X files contain the reference output intent with which they should be processed, so as to deliver a single job with everything embedded. In any PDF/X3 file (or a more recent version), all of the elements that do not match the output intent’s colour space must be tagged with ICC profiles to ensure correct colour conversions.
This file format is now broadly supported by mainstream software applications, including Adobe Illustrator and InDesign, making it very easy to specify how a sign shop’s customers should supply their jobs in the first place. They can also be pre-flighted very easily to ensure they are acceptable before printing. The files should be exported to the most recent version of PDF/X available.
This approach to standardizing converts all elements to a single colour space by way of reference profile and supports spot colours. The use of device-link profiles becomes unambiguous, as there is only one colour space after the process of standardization.
Indeed, the use of device-link profiles is far superior to typical ICC conversions. By allowing a workflow to calculate a direct link, it delivers the following advantages:
When the reference space is fully known, so too is the grey axis, for which the device-link profile can then be optimized.
Device-link profiles can maintain detailed midtones and shadows while saving significant quantities of ink. In addition to reducing overall costs, this extends the longevity of printheads, improves printability on various types of media and saves energy by enabling printing at lower temperatures.
The device-link profile contains information about which colour targets it needs to hit and how to direct a specific printer to return an image to its original state.
In practice, this means when a printer’s colours start to ‘drift,’ a simple print-and-measure process—with a relatively small colour target, involving approximately 500 patches for a CMYK printer, read with a spectrophotometer—can compensate. This means the printer will produce the same output result, day after day, without much external effort.
Keeping as much colour as possible
There is a common misconception that using a reference space will necessarily limit the capabilities of the printer to the common colour spaces shared by that reference and the printer’s own gamut. This is not always true, however, as there are several strategies that can also be implemented after a file passes through the standardization process:
Spot colour mapping
Spot colours will be mapped to the printer’s gamut and not be limited by the reference output intent. A job using GRACoL as the reference and two Pantone colours, for example, will actually be rendered as a six-channel job, i.e. CMYK plus two spot colours. This method ensures all blending and transparency effects—like drop shadows and overprints—are processed correctly, while enabling the RIP to address the full gamut of the printer.
The reference colour space—usually CMYK—can be mapped to the printer gamut in any way that is preferred. In some cases, for example, ‘punchy’ colours may be prioritized over printer-to-printer consistency or matching to a contract proof.
If the printer’s gamut is broader than the reference space, then perceptual mapping with black (K) compensation can expand the reference space and maximize the use of the printer’s gamut. If, on the other hand, the printer’s gamut is smaller than the reference space, then a hybrid approach of perceptual and colorimetric mapping is recommended. By keeping the colorimetric mapping within the achievable ‘middle part’ of the gamut and using the perceptual mapping to compress the edges of the gamut, a visually comparable image can be retained without flattening the edge colours.
For fine art reproductions and photography, an RGB reference space may be used, so as to deliver a broad gamut without separating the image to CMYK, where there could otherwise be breaks in smooth transitions.
After this type of standardization, the image can be mapped to the printer gamut. This way, the image benefits from both standardization and device linking.
A better way
The advantages of standardized printing are significant for the production of wide-format graphics. The key is to choose a RIP that is capable of (a) supporting the standardized printing process for inkjet devices—without the limitations most RIPs face due to their approach to standardizing colour management—and (b) using device-link profiles. The rewards are worth the investment.
Bart Fret is director of large-format sales for GMG Americas, which develops colour management software. For more information, visit www.gmgcolor.com.
Source URL: https://www.signmedia.ca/wide-format-printing-optimizing-colour-management/
Copyright ©2022 Sign Media Canada unless otherwise noted.