DIGITAL PROOFING -- Taking Measure of a ProofAugust 2003
As for how the rules have changed, Zwang reports the study found that color has become less of an issue. "Most of the color proofing vendors have adopted an ICC-based, color-managed workflow within their systems. As a consequence, if you look at the delta errors in IPA's test results, all of the systems did well matching to the targets, especially the known target. They averaged 2.0 to 2.5 delta E variance," he says.
There is a more important issue than color, which is proper processing of the files themselves, Zwang asserts. This is a legacy of the desktop publishing revolution.
"Output expectations now are being set upstream, where the file is created, and are being based on a file processed through a local RIP," Zwang points out. "Therefore, consistency between RIPs in file processing has become critical."
It's an up-hill battle, though. The consultant notes that a recent proofing study ("Proofing: The Customer Is Always Right") conducted by TrendWatch Graphic Arts found that almost 50 percent of the companies sampled had more than five color printers or proofing devices in their shops. And most of the systems are not the same.
"If the RIPs are different, you get different results. So which one's right?" Zwang asks. "Processing variability was one of the things we saw most in the proofing study. The handling of overprints, in particular, was all over the place. The Altona suite deserves a lot of credit. It tests everything, including overprints, spot color handling and color identification," he says.
"There was an underlying message to the testing that the proofing vendors didn't understand going in, but they heard it loud and clear coming out of the event: there is more than color to proofing, and it is a significant issue," the consultant explains.
Over time, a lot of the RIP vendors have developed solutions to work around inadequacies in the core RIP they get from a licensor, Zwang says. "There is still a lot of legacy equipment being used in the industry. That increases the need to standardize how file elements are handled, first off by making sure the core RIP developers design their solutions so they process files the same way."
Even though extensive testing already had been done in Europe, Zwang says there are several reasons why it was worth doing similar testing in the United States. "For one, we wanted to address the illusion on the part of some vendors that proofing requirements are somehow different in Europe from the U.S. Secondly, we needed to educate the users."
One participant came in with proofs from his five proofing systems, the consultant recounts. "He said, 'This has been a fascinating experience for me.' I asked why. 'All five of them are different. Which one is right?' "
Adding a blind exchange was another reason for repeating the test, Zwang says. "We proved that even with a blind exchange you can get amazing results, which validated the whole concept of exchanges using ICC," he asserts.
While the vendor sample included makers of virtual/on-screen proofing solutions, those systems weren't included in the test. Zwang didn't think they should be included for two reasons. "Even just dealing with hard copy, we knew the issues were going to be significant and confusing. Throwing those systems into the mix would add noise and confuse people even more," he explains. "Also, in my mind I had a question as to how to validate that the systems were accurately measuring and matching what they said they were."
Soft proofing vendors have created measurement tools, the consultant acknowledges, but he's not sure if they work across systems. "I'm not saying the tools are bad, because they could be very good. I just wasn't ready to jump into that type of testing. We might include them next time," he adds.
The Graphic Arts Technical Foundation (GATF) did evaluate soft proofing systems in its original research study. The Pittsburgh-based organization reports finding that the quality of monitor manufacturing "has matured to a point of consistency between devices that enables implementation of soft proofing with confidence in its accuracy." The "GATF 2003 Soft Proofing Study" examines both soft proofing for content (position, spelling, etc.) and contract color.
For a session at Tech Alert 2003, Greg Bassinger, GATF's manager of process controls, supplemented the study with a review of products from two leading vendors and a "home grown" system he put together from off-the-shelf monitors and software. Bassinger wanted to see how closely such a system could emulate the capability of "packaged" offerings.
Experts Weigh In
More than three dozen "color experts" compared images on each vendor's monitor to a printed sheet produced to GRACoL standards. The monitor-to-monitor results were said to be extremely close, but the home-grown alternative was noticeably inferior.
According to Bassinger, any study of soft proofing is currently limited by the fact that there is no method available for reading and calculating the differences between emitted and reflective colors, which would be required for a monitor-to-press sheet comparison. The primary reason is that the press color gamut exceeds that of a monitor.
One interesting finding of the study was that, when asked if a monitor image was a satisfactory "contract" predictor of the press sheet for four different sample images, at least half of the experts said yes. The affirmative responses reportedly reached as high as 85 percent for one image.
Soft proofing technology recently achieved a milestone, of sorts. SWOP (Specification for Web Offset Publications) certification has been a recognized measure of proofing quality for some time. The organization has now certified two on-screen/virtual proofing solutions.
George Leyda, president of Color-Info Technology, in Mahtomedi, MN, is in charge of the certification process. "We require that a measurement tool be provided for measuring the color on the monitor," he reports. "This solution may not equate to print measurements, but it should relate to itself.
"In our normal certification process, we include a double-blind evaluation by at least three 'industry color experts.' The methodology I've come up with (for soft proofs) is to do several hard-copy proof evaluations before we do the evaluation of a screen proof. Doing this type of pre-evaluation of other proofs gets the experts into the frame of mind of looking for 'color balance' and acceptability at the press in a real 'production' situation," Leyda notes.
Soft proofer certification then is based on an examination of the same numbers considered for any other proof, he continues. The key (for vendors) is to achieve an acceptable visual rendition and have the objective data agree, according to Leyda.
Another mechanism for evaluating proofers is on the horizon. Following the lead of SWOP, the organization behind GRACoL (General Requirements for Application in Commercial Offset Lithography) plans to develop a proofing certification program based on that standard, reports David Hultin, director of print initiatives at IDEAlliance in Alexandria, VA. "We are awaiting the official release from CGATS of TR004," Hultin says.
Pre-release versions of GRACoL Press Sheets have already been made available. The sheets reportedly can be used to evaluate whether proofing systems are reasonable predictors of the final printed product for commercial offset lithography. They include two files from the Altona Test Suite as prepared for the proposed GRACoL Grade #1 characterization.
The "Altona-Visual" page can identify correct processing of such elements as fonts, overprints, spot colors, RGB-CMYK conversion and shading. For color managed proofing systems, the sheet can be utilized by first measuring the ECI2002R target in the "Altona-Measure" page and then creating an ICC profile that is used as a target.
Pre-release GRACoL Press Sheets are available for purchase from IPA at www.ipa.org.
Controlling the Remote
Soft, or on-screen, proofing—even in a local, controlled environment—has only met with modest success to date. Limitations of existing display hardware and calibration technology are central to the overall problem. To be effective, virtual proofing implementations must address these concerns.
Current RGB display technologies fall broadly under two categories—CRT and LCD. Neither display type has the full range of properties required to enable critical-color, virtual proofing in uncontrolled lighting conditions.
CRT (cathode ray tube)
>> Pros: A mature technology, with sufficient color gamut, stable color properties, good uniformity across the monitor (top/down/left/right), smooth reproduction behavior and little change in color appearance when images are viewed from an angle.
<< Cons: CRT monitors have limited brightness, offer a maximum resolution of ~100 dpi, are difficult to correlate numerically with measured reflective color on paper, and are big and heavy.
LCD (liquid crystal display)
>> Pros: Monitors are capable of much higher intensity (up to three times the brightness of CRTs), offer higher contrast ratios (400:1 or more versus 200:1), utilize an inherently subtractive process like CYMK colors on reflective media, produce an acceptable gamut for simulating CMYK and provide up to 200 dpi resolution.
<< Cons: A less mature technology, typically with marginal screen uniformity (top/down/left/right), less smooth and more complex color reproduction behavior, and significant shift in color appearance when images are viewed from angles other than directly in front.
Kodak Polychrome Graphics (KPG) elected to base the initial version of its Matchprint Virtual Proofing System on CRT technology. However, recent improvements in LCD technology may enable a comparable LCD-based system as a next-generation product.
KPG's laboratory evaluation of existing high-end, calibrated CRT monitor systems indicated significant variation from the color target and from display to display. Although this error is considered numerically small by some standards, the company's testing showed a visual difference actually becomes noticeable to experienced color viewers within that range.
A true virtual proofing system requires implementing a calibration method for RGB displays to ensure consistency between any two monitors. Achieving an accurately calibrated system requires use of a precision measurement device. The requirement for the emissive color measurement device used is a repeatability of ±0.1 delta E and an absolute accuracy (versus the primary standard) of ±0.5 delta E for all color values. However, most low-cost emissive colorimeters and spectrophotometers commonly vary from each other by ±2 delta E or more.
Since virtual proofing sometimes can involve a side-by-side visual confirmation of colors displayed on the screen against those of contract CMYK hard copy proofs, a proper viewing environment is required. Light booths historically have exhibited variability with regard to color temperature, even among products claiming a standard viewing environment such as D50.
When a CRT display is used adjacent to a light box viewer, even slight differences between the systems become readily apparent. This problem can be addressed by selecting viewing equipment that:
* has uniform intensity, particularly from top to bottom; and
* contains light bulbs or fluorescent tubes that remain consistent over time and from lot to lot. (Commercially available D50 fluorescent tubes from different manufacturers—or even different lots from the same supplier—can vary significantly from the design center point of 5,000 K.)
Additionally, the illumination level in the hard copy viewing equipment must be reduced somewhat to yield an overall perceived brightness comparable to the CRT display device.
To ensure that stray light doesn't compromise the color, appearance and contrast of CRT images, a viewing environment that eliminates, or nearly eliminates, ambient light (e.g., a kiosk enclosed by an opaque curtain or walls) is necessary. It is possible that LCDs (by virtue of their significantly higher brightness) could be employed in a virtual proofing system with significantly less restriction on ambient light levels.
One of the challenges in remote proofing is getting remote devices to match each other or a common standard. Maintaining a satisfactory match between multiple proofing systems requires enforcing periodic calibration. This is why each virtual proofing site should have a designated color administrator and a system administrator (who can be the same person) to maintain the system.
A final quality assurance mechanism required with the use of a virtual proofing system is allowing for adequate warm-up time. Laboratory studies have shown that CRTs often emit more light when first switched on, but equilibriate and remain very stable for long periods of time once temperature stability is achieved. The user should install a local start-up application that detects whether the system has been shut down and rebooted.
This article was adapted from a Kodak Polychrome Graphics white paper titled: "Matchprint Virtual Proof: Technical White Paper." For an electronic copy (PDF), visit www.kpgraphics.com/info/DigitalLibrary.