 | |||
|
Vision Register Versus. 3-point Register Using Vision-Based Systems As A Quality Control Tool As implemented in NELA VCP and H-VCP, vision-based punch/bender is a system for correcting the accumulated register error in exposed printing plates. The system includes a pair of Charged Couple Device (CCD) video cameras feeding images of the register patterns to a Windows XP workstation equipped with special software and hardware, which analyzes the patterns and compares them to a set of master patterns stored in memory. The software makes a computational analysis about the sameness of the actual pattern sent in by the video camera compared to the master pattern. This analysis results in an interpretation of the differences between the two patterns. These interpretations are then converted to movement parameters, which are sent to the, Motion Control Table. Lateral, circumferential, and skew movements, also known as X, Y, and Theta (rotation) represent the required corrections based on the Cartesian coordinate system. X represents is the lateral axis, Y represents the vertical axis (circumferential), and Theta represents skew. The motion control table, which is holding the plate via a vacuum plenum, makes the moves that were ordered by the software. The entire process is repeated: cameras check the position, software makes the analysis and orders the changes, and the motion control table carries out the movements. This process repeats until the plate is exactly in the right position Vision notching and bending is therefore a closed-loop system that absolutely assures that only high quality plates are being delivered to the pressroom. Electronic or mechanical 3-point register "should" work, but…it is an open loop system. The plate touches the three points and the plate is punched and bent. If there are problems with the position of the image on the plate these problems will be passed on, and will not be discovered until they appear on the press. With vision, register accuracy is guaranteed…plates that cannot be brought into register by the vision system will not be punched or bent and must be remade. This quality control check prevents out-of-register plates from reaching the pressroom. The 5, 50, 500-dollar rule illustrates the benefit of the vision closed-loop system. The rule says that it cost five dollars to fix a problem discovered in composition. It costs fifty dollars to fix a problem discovered once the plate has been made, and it cost five hundred dollars to fix a problem when it is discovered on press. With non-vision systems, problems are only discovered on press (the five hundred-dollar kind of mistake). With vision the problems are discovered in pre-press (the fifty-dollar kind of mistake). Reinforcing this point is a quote from Mr. Dan Lay of C. J. Krehbiel Co, at the Fall 2000 Research & Engineering Council's Press Seminar. "A good press plate may cost $20.00, but a bad plate can cost $1,000 in lost production and rework." Vision-based notching and bending will deliver accuracy of 1/10th of a row of dots, correcting for burrs and other deformities on the edge of the plate, while non-vision based-bending (3-point) will deliver an accuracy of a little less than one row of dots. Both cases assume a 100 line per inch screen ruling. While both 'sound' pretty good, it is important to remember that register error is cumulative in nature. The accuracy as measured in the final product is the total accumulation of error that builds-up in the pre-press and pressroom processes. Even after controlling for all the possible pre-press error with a vision-based register system the plate still has to be mounted on the press by a pressman. Some error is introduced in the mounting stage, especially on slot lock-up presses, which have no mechanical device to tension the plates or hold them in place on the plate cylinder. These presses, which offer very fast makeready depend on the rotation of the press and the spring-like characteristics of the aluminum printing plate to keep them in place. Finally, the press has to be properly calibrated and some error is introduced here, since it is impossible to make every plate cylinder, exactly match every other plate cylinder. Therefore, it makes sense to deliver plates to the pressroom that have the smallest possible amount of error so that the cumulative error is as small as possible. Results in the field support this argument. Actual newspaper industry experience in Europe as reported by a major CTP manufacturer's Managing Director (President) is as follows: "With vision punching and bending, the bend-to-image accuracy is 20 to 40 microns (0.0007 to 0.0014 inches), without vision the accuracy is 50 to 150 microns (0.002 to 0.006 inches)." The result: on its best day 3-point is not as good as vision is on its worst day. With vision there is no question about where the problem is. If the plate cannot be brought into register and is rejected by the vision-based punch/bender, the problem is upstream. If the plate is accepted by the vision-based punch/bender, therefore notched and bent, and goes on press but is still out of register, the problem is downstream. Without vision there is no easy way to prove where the problem is. Problems can go on for a long time with no resolution.
FAQs
NELA * 610 Whitetail Boulevard * River Falls, WI 54022, USA
|
|
|
Further information about NELA products click here. 