Presented at the TAPPI Symposium Proceedings on Process and Product Quality, October 29-31, 1991, Rochester, COLOR COTROL FOR PROCESS COLOR PRITIG: EEBALL VS. SPC Robert Chung* Abstract A unique press run was conducted with the use of the RIT s M1OOOB Harris four-color web offset press. The objective of the press run is to verify the usefulness of color measurement and SPC as tools for color control during high-speed printing. In this press run, one-half of the press signature was controlled by a SPC team using a scanning color densitometer and its associated SPC software packages. The other half of the press signature was controlled by an eyeball team, i.e., only visual perception was used as the sensor for detection and correction of assignablecaused color variation. During the experiment, color variations were introduced to the running press unannounced and each color control team was challenged to restore the printing conditions back to normal. This paper discusses the experiment as well as the pros and cons of visual perception and instrument measurement for color control during process color printing. Keyword Process control Process capability Color printing Perception Color measurement Introduction To prepare for a major workshop during the RIT s Fifth Annual Conference on Quality and Productivity in the Printing Industry in June 1991, a pilot press run was implemented in April using the four-color web offset perfecting press located on the RIT campus. The workshop objective was to demonstrate and compare two different methods of press control for conformance to printing specifications. The rationale behind the workshop are as follow: (1) To experience the value of measurement and SPC in the Look-Measure- Think-Act approach to process control in printing; (2) To provide an argument that SPC can assist printers in controlling the press; (3) To encourage active participation and role-play from conference attendees. Here, participants will be asked to make press control decisions during a live web press run. For the pilot press run, a class of graduate printing students from RIT s School of Printing Management and Sciences were participated in the event. Methodology Design of Test Form A test form or signature was designed with two identical halves. Each half of the signature contains pictures, color control bars, 3-C gray bars, registration targets, etc. *Rochester Institute of Technology, Rochester,, USA; email:rycppr@rit.edu
p. 2 22.375" 35" Ink Pick-up Zone (3 - C Gray) Color Control Bars Color Control Bars press is running. Eyeballs and experience are the only tools for making color judgment. They are to provide directions to a designated press crew for color adjustment on one-half of the web. Figure 2. offers a simplified algorithm for detecting and correcting color variation by eyeballing and a work flow in the workshop. Color Image Reproduction Eyeball SPC (Operator side) Header & Sampling Coding Image Header & Sampling Coding Image Eyeball (Gear Side) Figure 1. The press sheet layout. Materials, Equipment, and Printing Conditions Press: Harris M1OOOB four-color perfecting Press speed: Make-ready at 400 fpm; Running at 1200 fpm Plate: 3M Viking GM Ink: Flint Ink down sequence: K, C, M, Blanket: Reeves, Vulcan 2000 plus 3 ply compressible Paper: S. D. Warren Somerset gloss, 50 lbs Width: 35 (24 ink keys) Cut-off length: 22.375 Sheet size: 17.5 x 11.188 41,469 ft / web roll 22,240 impressions / web roll Fountain solution: Rosos KSP #500 ASM4 SWOP Specifications: Solid ink densities based on the SWOP Hi-Lo target Two Independent Press Control Teams Color Image Reproduction SPC Eyeball team: Three members make up the eyeball team. The team will be supplied with press sheet samples at one-minute intervals when the Take corrective actions Start Obtain the OK press sheet Observe any randomcaused color variation Take a sample Compare sample with the OK sheet Is perceived color difference significant? Another sample? End Figure 2. Process Color Printing and Control by Eyeball. SPC team: Three members make up this team as well. Press sheet samples will be supplied to them similarly. Measurements will be made and SPC data displayed with the use of the GMS/Cosar AutoSmart scanning densitometer. They are to follow a decision-making path as outlined in Figure 3 to provide directions to a designated press crew for color adjustment on the other half of the web. Guidelines for instructing the pressman for inking adjustment include: (1) Indicate the treatment letter and the latest sample number received on an index card supplied, (2) Indicate the ink, either increase or decrease, and the
p. 3 amount in # of LEDs, and (3) This process should continue until no more ink change is required. At this point, the index card should be marked, o more change. Guidelines for introducing assignable-caused variations include: (1) One assignable-caused variation will be introduced per treatment, (2) The variation will be introduced globally, i.e., evenly across the press sheet, and (3) The maximum inking change will be limited to five LEDs and the minimum inking change will be at least one LED in either direction as displayed by the remote inking control console (Figure 4). Start Establish center lines & control limits for C, M,, K solids Take corrective actions Figure 3. Process Color Printing and Control by SPC. Troublemaker Take a sample Measure solid ink densities C ok? M ok? ok? K ok? Another sample? End Assignable-caused variation will be introduced to the press unannounced by the troublemaker. In this case, this author served the role of troublemaker. The instruction for unannounced press changes will be communicated directly to a designated color adjuster in the web press team. Figure 4. An illustration of inking profile across the web. Possible changes to the press inking settings are outlined below: (1) Varying one ink, i.e., either increase one of the four inks or decrease one of the four inks; (2) Varying two inks, i.e., either increase or decrease two of the four inks, or increase one and decrease one of the four inks; (3) Varying three inks, i.e., either increase or decrease three of the four inks, increase two and decrease one of the four inks, decrease two and increase one of the four inks, etc. In addition to inking variations, image-toimage registration variation will also be introduced unannounced to make sure that visual assessment of printed reproduction is not overlooked by the SPC team. The Plot Tasks of the web press crew are to perform press make-ready, to keep the paper roll going and delivery end cleared, and to keep image-to-image registration in control. A pressman was designated as the inking adjustment person who will implement unannounced press changes and work with both press control teams to implement the changes as instructed. The following portrays a sequence of events in the press run.
p. 4 Press Make-ready to SWOP Specifications The web press team reduces the ink sweep to minimize the effect of the LED change in ink feed. Then the press is adjusted to meet SWOP printing specifications 1. At this point, two halves should be the same in appearance and in density measurement. If there is a conflict between the two criteria, the former criterion takes precedence. Process Capability Study Both the Eyeball and the SPC team assess process capability of the web offset press whereby no intended change is introduced to the press. At least 30 samples will be collected at one-minute intervals for measurement or observation. Play and Learn Both the Eyeball and the SPC team learn the cause-and-effect relationship between changes in inking settings and visual detection or density differences in press sheets.about 90 minutes will be allotted for these activities. Process Characteristics Process Characteristics A. Linear B. on-linear Control Setting Figure 5. depicts two likely outcomes between the press control settings and the observed process responses. Color Control on Press Experiment An inking variation will be introduced to the press by the troublemaker as described. Each color control team uses its own detection, correction, and verification method to bring its side of the process back in control. All inking adjustments will be communicated in a written fashion. Once a team has reached its goal, it should be communicated to the web press team. An additional 3-5 minutes press running time will be allowed for the other, team to come to its completion. This, then, signifies the end of a treatment. Several such treatments will be introduced as time and resources allow. Print Analysis & Evaluation Two scanning densitometers were used to provide the after-the-fact print analysis on both halves of the web. The result of the press run were compared in terms of the accuracy and the consistency relative to the OK press, sheet. Disclaimers The designed press run is not a typical press run seen in the printing industry today. It differs in many ways: The layout of the signature includes heavy ink coverage with test targets, color images, and ink pickup zones. The eyeball team will assess color images and a 3-color overprint tint for color control. The SPC will measure color control bars. Control Setting Figure 5. Cause-and-Effect Relationship
p. 5 Two color control methods are carried out simultaneously. We therefore assume that there is no real difference between the gear-side and the operator-side of the printing units. We will introduce assignable-caused variations to the press globally to observe how each control method carries out its monitor/detect/correct/verify cycle. Sampling is very frequent at one-minute intervals. The metric for press adjustment is not the gut feeling of the pressman, but the number of LEDs of the ink feed indicator. Results Detailed planning and various organizational meetings held among faculty, staffs, industry representatives, and graduate students contributed to the overall success of the press run. The sampling rate of one minute worked well in the experiment. Equally significant was the usefulness of using LED to communicate relative inking changes between the pressman and both color control teams. The Table 1. highlights natures of the troubles and reactions of each color control team in LEDs. First of all, there were five treatments, labeled from treatment C to G, introduced in the press run. These treatments were either as one-ink or two-ink variation. Ideally, only five corrections are necessary. However, the number of times that the eyeball made corrections of are 12 and the SPC team made nine corrections. For the five treatments introduced, the eyeball team reacted to press changes faster but less precise. The SPC team was able to declare the completion of almost all treatments with less number of times for corrections. A bonus point that went to the eyeball team was their ability to detect registration change while the SPC team was busy making density measurements. Treat C M K Credit Remark C6-5 Troubmaker (1) One-ink var. C7-2 -2 4 Eyeball (1) 6 SPC (1) C10 +2 +1 Eyeball (2) C15 OK - SPC D3 +5-4 Troubmaker (2) Two-ink var. D4 Paper splice D5 +5 Eyeball (3) D6-6 +6 SPC (2) D9 +2 SPC (3) D11-2 -2 Eyeball (4) D12 OK - SPC D14-2 Eyeball (5) E3 +3 Troubmaker (3) Regis C-2 rows E4 Eyeball Out-of-regis. E5-3 -3 Eyeball (6) E7 +1 SPC (4) E8-4 Eyeball (7) E8-2 SPC (5) E9 Paper splice E11-2 SPC (6) E11-2 Eyeball (8) E15 OK-Both teams F6-2 -5 Troubmaker (4) Two-ink var. F7 +5 SPC (7) F7-4 +4 Eyeball (9) F8 +5 SPC (8) F11 +2 Eyeball (10) F12-3 SPC (9) F13 +1 Eyeball (11) F15 OK - SPC G1 Troubmaker (5) o change G9-2 -2-2 Eyeball (12) G10 OK - SPC Table 1. Records of interactions between troublemaking and corrective actions. Once the above data table was compiled and discussed in the class, there was a clear consensus that the SPC method had demonstrated a clear advantage as a preferred color control method than the eyeball method.
p. 6 Discussions Press Speed and Sampling Considerations Because of the fixed cut-off length of the plate cylinder, the time interval (in seconds) per impression can be I calculated from the linear press speed (in feet per ~ minute) as shown in Table 2. (sec/imp) = or (sec/imp) = X ft min x 1 0.0089 X 1 min 60 sec Press speed t/imp # of imp/min (fpm) (sec.) 400 0.28 211 600 0.19 317 800 0.14 422 1000 0.11 528 1200 0.09 634 1400 0.08 739 1600 0.07 845 1800 0.06 950 2000 0.06 1056 Table 2. Relationship between press speed and number of impressions per minute. When the press is running at the speed of 1200 fpm, a total of 634 press sheets are printed every minute. Since only one sample was pulled per minute, the number of samples relative to the entire press run is less than 0.2% of its population. et this is very much the limit of how much one can sample in real time with the state-of-the art scanning equipment available today. Other factors which influence how sampling was determined are: (1) Running time per treatment, (2) Time required to pick up and mark samples, (3) Time required to travel from the press site to the measurement station, (4) Measurement time per press sheet, and (5) Time needed to assess the measurement and control charts for possible corrective actions. 1 x 12 in 1 ft x 1 imp 22.375 in, where X is press speed in ft/min Assignable-caused Variations and Process Tempering One reason which increased the number of corrective actions for both teams is the roll change or paper splice during the middle of a treatment. This was evident in the treatments D and the treatment E. It was noticed that both density and color balance were affected right after the roll change. In other words, additional corrections were necessary to restore changes caused by roll changes, but not by the designated troublemaker. Two reasons which caused the eyeball team to make more moves than the SPC team are dot doubling/slur and process tempering. When the eyeball team examined three-color overprint tints from the press sheet, shifts in color balance could be caused either by the change in solid ink density or by dot slurring during the press run. Dot doubling/slur tend to come and go without any warning or prediction. In other words, any change in the solid ink density to attempt to restore color shift which was caused by dot doubling/slur can only create greater color variation. Process tempering by the eyeball team was evidenced in the treatment G. This was caused by the inability of the eyeball method to differentiate between perceptibility and out-of-control situation. The measurement of solid ink density and the use of SPC method by the SPC team avoided both flaws in the press run. The Role of Eyeball vs. SPC in a Press Run Figure 6. shows the number of steps a press run has to go through from beginning to end. It also illustrates the role of eyeball and instrumentation in verifying and controlling the printing process.
p. 7 Adjustment Loop o Start of job Press make-ready Sample okay? es Use visual inspection and in-house standards Judged subjectively by customer or his designate One could argue the necessity of Measure in controlling the press. One could also argue the danger of not having Measure as part of the process control method. But no one will deny the importance of Looking at the press sheet before anything else could happen. Both eyeball and measurement are used in the process. They serve different roles and need to work hand in hand. Process Control Loop Sign-off Press run Establish references for the press run Look Measure Think Act Think before Act ensures that we don t jump to conclusions before we understand what we are trying to accomplish. And statistical process control (SPC) is the tool which we help us Think. A Difference in Learning Curves End of job Figure 6. Printing Process Flow Chart The initial phase of the press makeready involves printing press sheets to either meet pressmen s own requirements or an in-house standard. The second phase of the press makeready is to seek the customer s sign-off by making necessary fine-tuning on the press. Because of the subjective nature of visual judgment, or the need to match the press sheet to a color proof, the color adjustment loop can be lengthy and costly. Once the press sheet is signed off, good copies are collected. The rest of the challenge is to maintain the ink-water balance of the press to produce the remaining number of copies as ordered. To ensure that the press is in control and is producing good copies, samples are pulled and contents evaluated. Here, the proposed color control method is Look-Measure-Think-Act. We believe that there is a learning curve for either method of color control. The eyeball method has worked for years in the printing industry yet it heavily depends on the ability and the skills of the individual. We also believe that proper measurement, when coupled with proper SPC techniques, will provide printers with a scientific approach for process control and defect prevention, thus, to obtain a higher degree of conformance to customer's requirements (Figure 7.). Conformance to specifications Time Figure 7. Expected Learning Curves. SPC Eyeball
p. 8 otice that the point of the cross-over, as shown in Figure 7, is unknown and is believed to be different from printing company to printing company. But there are ample examples of such cross-overs in other industries, e.g., automobile, computer, etc. Literature Cited 1. SWOP Inc., SWOP, Recommended Specifications Web Offset Publications, 1988 Ed.,,. Acknowledgments The author wishes to thank the following groups of people from RIT who made this project possible: Students in his Tone and Color Analysis class in the School of Printing Management and Sciences, T&E Center, and the Laboratory for Quality and Productivity in the Graphic Arts. He also wishes to acknowledge the help and participation of the following companies: GMS Cosar, Tobias Associates, and DuPont Imaging Systems.