It seems only yesterday that debate raged, or at least quietly burbled, regarding plate curves and color bars. “Should color bars be curved, or should they be linear?” was the question that inquiring minds wanted to know. Happily, that question has been resolved for most, if not all printers. The absolute condition of the press, as represented by linear color bars, has been deemed to be far less important that the actual net color results as measured by color bars that have received the same curve treatment as the actual images, and the large majority of printers give their color bars exactly the same treatment as the rest of the job. After all, we want the color bars to be a true reflection of the actual appearance of the job.
But what about ink optimization? (AKA ink savings, GCR, or various sorts of secret sauce). Should color bars get the same treatment as the image, as they do with curves, or should they shielded from optizimation treatments? Based on informal conversations, it seems like quite a few users feel that indeed, ink optimization should, like curves, be applied globally, to image and color bars equally; but in fact, the situation is completely different, and ink optimization is not only inappropriate for application to color bars, but is absolutely detrimental to their utility and may even represent a sort of false representation of results, particularly in a G7 enviornment.
To illustrate exactly why optimization is so detrimental to the utility of color bars, I’d like to start with a basic color bar of the kind that is frequently configured to fit onto the end-flap of package products (and thus cannot easily escape the effects of optimization, since it is integrated into the image).
This little collection of just ten patches is just enough to judge compliance to G7 specifications with great accuracy, and can either be spread across the press sheet in the traditional manner, or as stated above, hidden into the end flap of the product. It includes solids of CMY and K as well as 25 50 and 75 percent patches of K and CMY grays, A perfect, minimal and effective QC tool. But will it withstand ink optimization?
To see, I have integrated it into some images we all know very well.
There they are, our three favorite girls, the SCID musicians, with a G7 color bar in the corner to assure that all is fine and in perfect gray balance (subject to the limits of web viewing).
In order to simulate a perfect storm, I have created a rather serious color imbalance, the kind that the G7 specification is designed to monitor and control. In this case, while solid ink values have remained perfectly constant, a shift has happenned in the Cyan and Magenta TVI, causing a shift in gray balance that is equally visible both in image and color bars.
The image here is seriously out of balance, but the color bars have done their job. The 25 50 75 CMY gray bars show the same shift we see in the image, allowing process control specialists to spot the unacceptable color variation and reject the job.
Now let’s take a look at the same imagebut this time including aggressing GCR/ink optimization.
This image looks like at first like a victory for ink optimization. This image of the musicians, which were subjected to the same imbalances as the unoptimized image, looks better than the unoptimized image. There is still shifting of the clothes, the flesh tones, the background, but the shifting has at least been reduced. But what about the color bars? Oh my! They are absolutely perfect! Any objective measurement of the color bars would show that we have a 100% perfect job, with complete compliance to G7 in every way, But there is one small problem. The actual image, while better than the unoptimized image is still visually unacceptable, yet based on the color bars, it would have passed QC inspection. How did this happen?
If we peel away the black to look at CMY and K separately, we see that black tints are reproduced with tints of black, amd CMY tints are reproduced with tints of CMY, just as expected.
Performing the same operation on the optimized file, we see that the black tints are reproduced using tints of black, and CMY tints are reproduced with….tints of black! They have completely lost their predictive power as QC devices, and no amount of TVI shifting will cause the slightest shift in gray balance.
With optimization, the image fares less well than the color bars. While GCR does an excellent job of protecting neutrals from shifting, this effect diminishes as we move away from the neutral axis to more saturated colors. Since most images contain many colors, they will invariably tend to fare worse than color bars designed with perfect CMY neutrals.
Does this mean we shouldn’t use ink optimization? Not at all. Optimization/GCR of images with non-optimized color bars is an excellent way to save ink, shorten makeready and increase stability on a wide range of images. But optimization is entirely inappropriate for color bars. Optimized G7 color bars can produce “false negatives” indicating a G7 compliant job even when TVI values and corresponding image appearance are seriously shifted. Be sure that across-the-sheet plate colorbars are excluded from the optimization portion of the workflow. And if colorbars are integrated into the image, such as patches hidden into packaging end flaps, be sure that any image optimization is done off-line, prior to plating, so that color patches can be excluded.