The Flemish Innovation Centre for Graphical Communication (VIGC) has released a study on the accuracy of spectrophotometers used within the printing industry.
“Color quality is the biggest challenge in the printing industry.”, says Eddy Hagen, managing director and trend watcher of VIGC. “Graphic arts companies will try everything to get the colors as desired by the customer. Those customers will use it as the most important criterion to accept, or reject, a print job. Which makes the devices to measure that color quality quite essential. So you would expect that the quality of those devices is top class. But it isn’t.”
If you manage color, click over to the study for the report: VIGC study on spectrophotometers reveals: instrument accuracy can be a nightmare
Discussion
By Henk Gianotten on Sep 15, 2008
The problem of inaccuracy and lack of inter-instrument agreement is a nightmare. However, it's documented by several research organizations. Photospectrometers need proper instruction, maintenance and proper recalibration.
The HDM in Stuttgart (what RIT is for the US, is HDM for Germany) did 2 research projects in 2004 and 2005. They checked 40 different devices from the attendees of the Digital Proofing Forum in 2005. The report was made available in German and English. The same year the University of Wuppertal (North-West Germany) arranged another Digital Proofing meeting to measure differences in hard- and softproofing systems and the equipement.
Stefan Brües and Claudio Höcker suggested to expand the Fogra Media Wedge (Fogra MediaWedge 3.0 is the result) and proposed to use the delta E2000. Not only in research tests but also as part of the measuring systems.
The Germans came to the same conclusions. Improvements are essential and the discussions how to measure and how to check the equipment are needed.
In my opinion, some problems mentioned by VIGC are international. We need to invest in the right equipment and have to spend on the proper training. Dirty lenses, calibration tiles and light sources! It sometimes seems that we maltreat these modern devices.
Funny, if one remembers the discussion in the early seventies on densitometers! Bad "white check" tiles, sloppy maintenance, old filters and so on.
The French say: "L'histoire se repète"!
By Erik Nikkanen on Sep 16, 2008
Many years ago on the forum "sci.engr.color" I read the discussions of the colour experts about the industries "dirty little secret" which was about the problem of the lack of colour instrument agreement. It seemed that it was common knowledge to them. Added to that were other comments that stated that normal spectrophotometers were just not capable of measuring accurate colour and instruments that measured at each wave length were required. Then the discussion went on to say that even these instruments were not good enough and that one would have to go to a national lab where very expensive instruments were used. Of course the costs of the instruments went up and up to the point of several hundred thousand dollars to have an instrument that was accurate.
So in hearing this amazing discussion, I came to conclude that the type of instruments one uses in the graphic arts will not be accurate no matter how much you calibrate them because they are not designed to be highly accurate for cost reasons.
The other half of the issue is that colour science is a highly mathematical technology but which is based on the average response of a group of people that were a part of tests back in 1931. Since it is an average of a group which I think was not that big and was mostly men, colour science is not what one would call an exact science. Many times, colour science is more at home in the psychology departments at universities than in physics or engineering departments.
Understanding the limitations of the science and the technologies to measure it, implies to me at least that one should not get too picky about colour accuracy. You will not get it always.
There are also two issues regarding the reproduction of colour. Accuracy and consistency. Instruments might not be accurate enough to ensure inter instrument agreement but they are very good to determine consistency in a production run.
Consistency is very important and in offset printing it is still a problem. It does not have to be but there is still no interest to design out the sources of variation that still exist in modern offset presses. Maybe the industry just loves to chase its tail.
By Paul Lindstrom on Sep 20, 2008
We at Digitaldots have made similar tests, but in our case (with fewer instruments involved), the difference between instruments was smaller, approx +/- 0.5 Delta E (1976).
But if one spectrophotometer had UV-filters, and another didn't, the measurements differed so much that you realize (or should realize), that something is wrong here.
We have also found, later, that if you measure a substrate when still in the area of the viewing booth, the results differ significantly than if you measure fx at a window, with stray daylight mixing with the light from the lamp in the measuring device. This may seem obvious, but may also be neglected by inexperienced operators.
So -- should we just forget about measuring the prints, and rely on visual evaluation only? Ofcourse not. As Erik Nikkanen concludes -- consistency is what we want, and for that we need to measure.
If we then test our colour discrimination capacity (color vision), we will find that the human eye (at best, if not "color blind") is very capable of detecting even small colour differences. We can then compare our measurements with what we see, and judge if we are accurate enough when we try and reproduce a certain colour.
Even if we still have a series of issues to solve in regard to practical "hands-on" colour management, the tools available today is much better than let's say 5 years ago.
And yes, at the moment there seem to be good reasons to start using the formula of Delta E 2000 instead of the older one, but some 5 years ahead we might have yet other, better options.
/Paul L
By Erik Nikkanen on Sep 20, 2008
"So — should we just forget about measuring the prints, and rely on visual evaluation only? Ofcourse not. As Erik Nikkanen concludes — consistency is what we want, and for that we need to measure."
The mistake the industry has always made is that for consistency one needs to measure. This thinking has lead the industry in the wrong direction. Measuring does not provide consistency. Process capability provides consistency. Measuring can be used to calibrate the process and follow it to confirm consistency. If the measuring is used to follow and then adjust the process, then it is used to correct errors in the process that does not have the process capability to be consistent.
The action of measuring does not prevent the variation. If one was really interested in consistency then one should design out the sources of variation and that would result in very little need for measurement.
I am not against measurements. It is very much needed but I protest when I hear that measurement provides consistency. The Printing industry has always got this thinking backwards.
By Wilco de Groot (IGT) on Sep 21, 2008
Garbage in = garbage out! Do not blame the spectrophotometer but look to the users and train them.
We have done a similar study, about 50 instruments, they do give the differences as mentioned in all comments before. However only very few where on one or two colors outside the tolerances specified in the practical standards for color (12647, 2846), none of them for the primary colors! Even in the dirty, bad calibrated (or not at all) state as we tested them.
As long as most users of the spectrophotometers have no idea of color theory or at least no proper instructions, L*a*b* specs are accepted by sales departments, cleaning practice is unspecified, confusion on backing is standard and many more small issues, I suggest to focus on training and information first and (probably much) later on making better instruments.