The Optimal Method for Press Calibration

No Favorite

Using data from a highly touted G7 pressrun, we showed that neither calibration technique was superior. The near neutral technique achieved better gray balance; the TVI technique produced better shadow rendition. We proposed a cooperative effort with Europe to develop an improved technique both factions could embrace. But by then, the G7 method was being heavily marketed, and our technical concerns were unwelcome, to put it mildly.

Both Idealliance and FOGRA (Germany’s Graphic Research Technology Foundation) had a financial stake in their preferred calibration technique, and there would be no compromise. We attended CGATS meetings for a few more years, but it became clear Idealliance was in control of US print standards. One of our last contributions to CGATS involved the definition of gray, which underpins the G7 method. At the CGATS meeting in Grand Rapids, MI in July 2009, we gave this presentation.

But once again, sound technical thinking was unwelcome. It was time for us to move on.

Opportunity Knocks

“One door closes; another opens,” as they say. By giving up on standards work, we were able to focus on finding a better solution to the calibration problem. We realized that standard datasets were more important than their associated press calibration technique.

ICC profiles built from these datasets were used to make color-managed proofs, and that is what printers wanted to match. It didn’t matter how these profiles were created, so long as they defined a realistic color space.

• In 2006, CGATS released a GRACoL dataset and two SWOP datasets, supposedly created using the G7 method
• FOGRA already had numerous dataset created with TVI curves, per ISO 12647-2
• In 2013, CGATS.21 was released, which included seven datasets for different papers, with similar appearance

By reframing the calibration problem as matching a reference profile, we could work with believers of either faith. Calibration is the adjustment of a device or process to produce accurate results. The Optimal Method is a general solution to the print calibration problem that works with any set of measurement samples. By using near neutral samples, it makes G7 curves. By using process color ramps as samples, it makes TVI-like curves.

Better yet, by using a full gamut of realistic color samples, it makes optimal curves. The Optimal Method works by reducing color errors to their minimum. This idea was inspired by the work of Robert Chung of RIT, who first used error statistics (CRF) as an indicator of color matching accuracy (see Figure 4). We presented The Optimal Method in yet another TAGA paper, at the 2018 annual meeting in Baltimore, MD. For details, click here.

Residual Error

Some engineering problems have no perfect solution. Building tone curves is that kind of problem. Ideally, tone curves would eliminate all color errors between actual printing and the reference profile. But this is generally impossible, and there will always be some residual error. The best one can do is reduce the residual error to a minimum, focusing on the errors that matter most. We call this the optimal solution, and that is what The Optimal Method finds.

We realize it may be difficult for those familiar with the TVI or G7 methods to accept this way of thinking. Both of those methods yield an exact solution for the sample sets they use. There is some comfort in thinking you’ve computed perfect curves down to the nth decimal place. But most of the possible colors are not even considered, relying on faith that if a few colors are correct, the rest will follow.

Those of us with experience running a color scanner will probably have a strong belief in the gray scale. And those with experience running a press may believe in TVI. But regardless of your faith, if you spend a little time using The Optimal Method (see Figure 5), you will see its advantages.

The Optimal Method, although simple in concept, is quite difficult to implement. So, we’ve created a free software tool called PressCal. Actually, we created a tool by that name in 2004, to compute curves for our TAGA paper. That tool gained acceptance over several years. In March 2010, I met with Dave Pauley, the CEO of Neyenesch Printers, in San Diego. We had a good conversation, where he encouraged me to turn PressCal into an open-source project. Today, the “color toolkit” is a library of 49 modules containing more than 50,000 lines of code. To view it, click here.

PressCal builds on the color toolkit and provides a full suite of calibration functions, including TVI, G7 and SCTV curves. It supports spot colors and expanded gamut (EG) processes. View the sell sheet.

We believe The Optimal Method has many benefits for flexographic printers. Unlike other calibration techniques, the color reference is an ICC profile. This could be a CGATS profile, a FOGRA profile or your own custom profile. If you use Flexographic Image Reproduction Specifications & Tolerances (FIRST) methodology, PressCal could save you pressruns. It can produce tone curves from the fingerprinting pressrun to match your target color space, process control aim points and characterization data of the calibrated process. PressCal can calibrate your press for re-runs or calibrate multiple presses in different locations to print identically, using your characterization profile. The latest version of PressCal has many new features to support flexographic users. Best of all, it’s free.

We believe print standards should be free, technically sound, and beneficial to both the industry and its customers. We’re not in the standards or software business. We won’t sell you a medallion for your website. But we are available as consultants to work with you on pressroom calibration.