Joel Zlotnick, Jordan Brough and Troy Eberhardt

This paper is the seventh in a series focusing on novel security design strategies that interrupt counterfeiting processes necessary for reverse engineering of document artwork.[1] The foundations of the strategy are the adoption of a monochromatic design gamut and the overprinting of multiple layers of translucent offset inks of the same colour to generate areas of higher image density, instead of by widening lines.[2][3] Split fountain printing was adapted for a new role in combating artwork reverse engineering.[4][5] The exploration of split fountain hardware variables started in Part 6[6] is continued in this Part 7 paper, where split fountain width is evaluated.

The papers in this series collectively describe a novel security design strategy that disrupts counterfeiters’ prepress artwork replication processes by preventing the isolation of individual offset plate images. Parts 6 through 8 of this series focus on hardware variables associated with split fountain printing, including split fountain placement across the roller width[6], split fountain transition width (this Part 7 paper) and splits between more than two pure inks (future paper, Part 8).

As stated in prior work, the ideas presented here are conceptual and will require adaptation and testing before finding application in real world printing environments. For simplicity, this work assumes that the opacities of several layers of translucent ink, when overprinted, add linearly to produce a darker image regardless of print order, the specific ink opacities combined, or ink layer thickness.

About split width
Though split fountain printing encompasses several hardware variables, the scope of this paper focuses only on split fountain transition width. Consider Figure 1 and Figure 2, which were captured at identical magnifications. In Figure 1, the pink to blue transition is abrupt and occurs over a short horizontal distance, while the transition from orange to grey in Figure 2 occurs over a wider horizontal distance. The difference is due to the split fountain configuration, including the degree of oscillator roller movement on press, which may be over a short distance (causing less colour blending, as in Figure 1) or a longer distance (causing more colour blending, as in Figure 2).

DE_21643_Fig01
Figure 1: Split fountain transition from pink to blue, occurring abruptly over a narrow horizontal distance in the centre of the image. Compare to Figure 2, which was captured at the same magnification.
DE_21643_Fig02
Figure 2: Split fountain transition from orange to grey, occurring gradually over a wide horizontal distance. Compare to Figure 1, which was captured at the same magnification.

Figures 1 and 2 illustrate the conventional split fountain blending of two ink colours. For this paper, the variable of split fountain transition width will be repurposed to simulate monochromatic, multiplate false split fountain effects that help conceal the true press setup. The examples presented below show the effects of changing the split fountain transition widths between ink stations. The first example (Figures 3 through 13) incorporates paired split fountains that can simulate three flat tones and eight false split fountains. The second example (Figures 14 through 20) does not use paired split fountains, and replaces the flat tones with three additional false split fountain effects.

Figure 3 shows the distribution of inks across the four ink stations in a hypothetical printing press, using only inks of the same colour but different opacities. The first station features a split fountain transition between pure inks of 8% opacity at the edges and 30% at the centre of the fountain, and so on for the other stations. The same two ink opacities (8% and 30%) are used in opposite placements in the first two fountains, and the same two ink opacities (18% and 44%) are used in opposite placements in the third and fourth fountains. These simplified ink opacities are contrived to make the examples easier to understand, and do not represent real ink formulations.

Print
Figure 3: Opacities of pure inks in each of four split fountain ink stations, expressed as various combinations of inks of 8%, 18%, 30% and 44% opacity. The first two fountains feature the same split fountain transition widths but opposite ink placements, as do the third and fourth fountains. The split fountain transition is more gradual in the third and fourth fountains than in the first and second.

Although all four rollers in Figure 3 contain a split fountain, the width of the transition between the pure inks is narrower in the first two fountains and wider in the second two fountains. The first two fountains in Figure 3 show wider swaths of the pure inks because the split fountain transition occupies less of the fountain width, and the third and fourth fountains show narrower swaths of the pure inks because the transition area is wider. The security value becomes apparent when multiple fountains are overlapped to simulate plates that do not exist in that actual press setup, as described below.

Simulating flat tones
Parts 5 and 6 of this series described the simulation of three flat tones by overprinting identical line artwork from two paired split fountain printing plates. The example is repeated to illustrate that flat tones can be simulated even in press configurations that vary split fountain transition width.

Print
Figure 4: Overlap of paired split fountain inks from Figure 3 to simulate flat tones. These flat tones depend on paired fountains in Figure 3 being configured with the same split fountain transition width, but opposing ink placements.

In Figure 3, the split fountain transition width is the same in the first two fountains and the same in the third and fourth fountains. Overprinting the first two fountains, or the third and fourth fountains, or all four fountains together produces one of the three flat tones shown in Figure 4, though every fountain in Figure 3 actually contains a split fountain.

Simulating dynamic transition speed
More interesting effects can be created by overlapping two split fountains that feature different transition widths. For example, both the first and third fountains of Figure 3 feature a lower opacity ink at the edges and a higher opacity ink in the middle, and both the second and fourth fountains feature a higher opacity ink at the edges and a lower opacity ink in the middle, but in each pair the transition widths are different. These combi­nations are shown in Figure 5, and simulate a dynamic opacity transition speed. The concept is explained in Figure 6 using just the first and third fountains.

Print
Figure 5: Overprinting of the first and third fountains, and the second and fourth fountains, from Figure 3. These combinations simulate an opacity transition that appears to progress more or less rapidly in different areas of the roller width.

In Figure 6, each pure ink occupies a portion of the roller width (marked with the letter ‘A’) where it has not been blended with the other ink. Part of each individual fountain width is also allocated to the split fountain transition (marked ‘B’), where the pure inks are blended and neither pure ink is visible by itself. In the first fountain, more of the width is occupied by the pure inks (‘A’) and the split fountain transitions (‘B’) are narrower. In the third fountain, less of the width of the roller is occupied by the pure inks (‘A’), leaving more room for wider split fountain transitions (‘B’).

Print
Figure 6: In each of the individual fountains 1 and 3, ‘A’ designates only pure ink and ‘B’ designates the split fountain transition area. In the combination 1+3, ‘C’ designates where both fountains print pure ink, ‘D’ designates both fountains printing split fountain transitions, and ‘E’ designates a pure ink overlapped with a split fountain. Compare to Figures 3 and 5.

Now consider the combined image in Figure 6, where the width can be divided into three areas. Areas where both the first and third fountains print pure inks (both ‘A’) are marked with the letter ‘C’ in the combined image. Areas where both the first and third fountains print split fountain transitions (both ‘B’) are marked with the letter ‘D’ in the combined image. Finally, areas where the third fountain prints its split fountain transition (third fountain ‘B’) and the first fountain prints its pure ink (first fountain ‘A’) are marked with the letter ‘E’ in the combined image.

In the combined image, pure ink areas marked ‘C’ are composed of two areas of pure ink from fountains one and three (‘A’), and show no opacity transition. The areas marked ‘D’ where both split fountains intersect are defined by the width of ‘B’ from the first fountain. The opacity changes more rapidly in the ‘D’ combination areas than in either of the split fountain areas marked ‘B’ because both split fountain opacity transitions are combining in an additive way. In the areas marked ‘E’, one split is overprinted with a pure ink, resulting in an opacity change that appears more gradual than in the ‘D’ areas where the two splits work together.

The overall visual effect is of no opacity transition (‘C’) to a gradual opacity transition driven by only one split fountain (‘E’) to a faster opacity transition driven by both split fountains moving in the same direction (‘D’). This effect is a simulation of different roller oscillation speeds at different points across the fountain width. As conventional split fountains use only a single oscillator roller with a constant movement, it would be challenging to replicate the subtle visual effect shown in Figure 6 using a single ink fountain without resorting to half­tones. A similar exercise could be done with the second and fourth fountains shown in Figure 5.

Print
Figure 7: Overprinting of the first and fourth fountains, and the second and third fountains, from Figure 3. The pairs do not simulate precise flat tones because the split fountain widths are not identical.

Other combinations featuring opacity transitions moving in opposite directions instead of the same direction are illustrated in Figure 7. These combinations include overprinting the first and fourth fountains, or the second and third. In Figure 7, an increase in opacity in one fountain tends to correspond with a decrease in opacity in the other, but due to the differences in split fountain transition widths, the result is not a perfect flat tone (as in Figure 4). Instead, subtle patterns of higher and lower opacities result. These will not be examined in detail here, but do contribute to the gamut of simulated patterns that can be derived from the actual press configuration shown in Figure 3.

Figure 8
Figure 8: Combinations of three of the fountains shown in Figure 3. Individually, each of these combinations just looks like a darker version of one of the split fountains shown in Figure 3.

Three-plate combinations
Figure 8 shows combinations of three of the four fountains shown in Figure 3. Any combination of three fountains from Figure 3 will include one of the flat tones from Figure 4, plus another one of the individual plates from Figure 3. The result is that the combinations shown in Figure 8 do not exhibit any of the unusual split fountain simulation effects shown in Figures 5 and 7, and just look like darker, lower-contrast versions of the individual fountains shown in Figure 3.

Artwork including flat tones
Figures 4 through 8 show combinations of two, three or four of the ink distributions shown in Figure 3, but do not address plate artwork. To conceal the actual press configuration shown in Figure 3, a security design could be constructed entirely or primarily from the fountain combinations shown in Figures 4 through 8. Such a design limits the ability of counterfeiters to inspect the individual fountains shown in Figure 3 in isolation, making it hard to inspect and segregate individual plate images, count the number of plates used, and/or redraw the individual plate artwork.

Figure 9: Split fountain artwork composed of inks of 8% opacity at the edges and 30% opacity in the middle, from the first station in Figure 3. The width of this split fountain is the same as the width of the split fountain shown in Figure 10, and it has opposite ink placements compared to Figure 10.
Figure 10 (plate2)
Figure 10: Split fountain artwork composed of inks of 30% opacity at the edges and 8% opacity in the middle, from the second station in Figure 3. The width of this split fountain is the same as the width of the split fountain shown in Figure 9, and it has opposite ink placements compared to Figure 9.
Figure 11 (plate3)
Figure 11: Split fountain artwork composed of inks of 18% opacity at the edges and 44% opacity in the middle, from the third station in Figure 3. The width of this split fountain is the same as the width of the split fountain shown in Figure 12, and it has opposite ink placements compared to Figure 12.

Some example plate artwork designs are shown in Figures 9 through 12, with each plate image inked from one of the split fountains shown in Figure 3. The composite artwork shown in Figure 13 is created primarily from overlaps of artwork on two, three or four plates, with artwork from a single plate displayed only in restricted areas to limit exploitation by counterfeiters. Several split fountain transitions are displayed across the width of the branches at the bottom of Figure 13. Most of the visible splits are simulations that do not exist in the actual press configuration, and several simulate changing opacity transition speed as described in Figure 6. Figure 13 contains diverse elements that include parts of four actual plate images (from Figure 3), three simulated flat tones (from Figure 4) and eight false split fountains (from Figures 5, 7 and 8).

Figure 12 (plate4)
Figure 12: Split fountain artwork composed of inks of 44% opacity at the edges and 18% opacity in the middle, from the fourth station in Figure 3. The width of this split fountain is the same as the width of the split fountain shown in Figure 11, and it has opposite ink placements compared to Figure 11.
DE_21643_Fig13
Figure 13: Composite artwork produced by overprinting Figures 9 through 12, including four actual split fountains, three simulated flat tones and eight simulated split fountains. The branches display several simulated split fountains, including some where the speed of opacity change appears variable.
Figure 14
Figure 14: A different split fountain configuration of the same inks shown in Figure 3. Unlike Figure 3, each station has its own unique split fountain transition width, so the paired fountains used to simulate flat tones are not present. The transition becomes progressively wider from the first fountain to the fourth. Compare to Figure 3.
Figure 15
Figure 15: Overlap of split fountain inks from Figure 14. These eleven false split fountains replace the combinations in Figures 3 through 8 when the individual fountains are configured as in Figure 14. Each combination features its own particular pattern of higher and lower simulated opacity transition speeds, as illustrated previously in Figure 6.

Unique split width in each fountain
The example illustrated in Figures 3 through 13 includes some fountains configured as opposing pairs, allowing simulation of the flat tones shown in Figure 4. Con­figuring the split fountains so that the width of each transition is unique prevents simulation of flat tones, but instead allows for three additional split fountain effects in place of the flat tones. As an alternative to Figure 3, which included opposing pairs, Figure 14 shows four fountains that each feature a custom split fountain transition width. For example, the opacity transition occurs over the shortest horizontal distance in the first fountain, and the longest horizontal distance in the fourth fountain. The fountains in Figure 14 can be overprinted to produce the eleven combinations shown in Figure 15, which facilitates two distinct security advantages.

Security advantages
First, designs created from the various overlaps in Figure 15 include a gamut of false split fountain effects that work to conceal the configuration of the actual split fountains used in the press (see Figure 14). This masks the actual plate count and inhibits reverse engineering of the true press setup by counterfeiters. Second, each of the combinations in Figure 15 features its own unique pattern of simulated split fountain transition speeds (similar to the one example detailed in Figure 6) where the rate of the split fountain opacity transition appears to change in different horizontal areas of the fountain. A counterfeiter attempting to simulate one of the effects in Figure 15 using a conventional approach in just one ink station could find the task challenging, because a single fountain can only accommodate a constant (fixed) oscillator roller movement. Though a counterfeiter could achieve such a simulation with halftones, counterfeit quality would be curtailed and microscopic detection of the counterfeit would be straightforward. A second artwork example, derived from the press configuration in Figure 14, is shown in Figures 16 through 20.

Figure 16
Figure 16: Split fountain artwork composed of inks of 8% opacity at the edges and 30% opacity in the middle, from the first station in Figure 14. The width of the split fountain transition is narrower than in the plates shown in Figures 17 through 19.
Figure 17
Figure 17: Split fountain artwork composed of inks of 30% opacity at the edges and 8% opacity in the middle, from the second station in Figure 14. The width of the split fountain transition is narrower than in the plates shown in Figures 18 and 19, and wider than the plate shown in Figure 16.
Figure 18
Figure 18: Split fountain artwork composed of inks of 18% opacity at the edges and 44% opacity in the middle, from the third station in Figure 14. The width of the split fountain transition is narrower than in the plate shown in Figure 19, and wider than the plates shown in Figures 16 and 17.
Figure 19
Figure 19: Split fountain artwork composed of inks of 44% opacity at the edges and 18% opacity in the middle, from the fourth station in Figure 14. The width of the split fountain transition is wider than in the plates shown in Figures 16 through 18.
Figure 20
Figure 20: Composite artwork produced by overprinting of Figures 16 through 19. Compare to Figure 13, with particular focus on the simulated and actual split fountain effects in the branches at the bottom of the design.

Conclusion
This paper explored the width of split fountain transi­tions as the second of three split fountain hardware variables that can facilitate security designs resistant to reverse engineering. Within the monochromatic design paradigm, varying split fountain transition widths between ink fountains can simulate the presence of extra press stations in printed artwork, and conceal the actual press configuration from counterfeiters. Possible visual effects can include the false split fountains described in this and prior papers, but also the simulation of varying oscillator roller speed across the width of the image, which could be difficult for even skilled counterfeiters to attack without resorting to halftones. The techniques described in this Part 7 paper could be combined with techniques introduced previously in Part 6 to optimise both the position and width of split fountains. Next, Part 8 of this series will demonstrate the inclusion of three pure inks in each split fountain.

Please note: The views expressed in this paper do not necessarily represent those of the US government.

MORE ABOUT THE AUTHORS

Joel Zlotnick is employed by the US Department of State, Bureau of Consular Affairs, Counterfeit Deterrence Laboratory as a physical scientist. He conducts research on how design strategies can help maximise the security value of printing technologies and security features, and develops training programmes on counterfeit detection.

Jordan Brough is employed by the Homeland Security Investigations Forensic Laboratory as a forensic document examiner, specialising in adversarial analysis and counterfeit deterrence. Jordan spends his time examining suspect documents and consulting with United States security document designers.

Troy Eberhardt is employed by the US Immigration & Customs Enforcement Homeland Security Investigations Forensic Laboratory. He supervises the Research and Development Section at the laboratory, which specialises in identifying and mitigating vulnerabilities within travel documents.

References

1 Zlotnick, J., Brough, J. and Eberhardt, T. (2016). Interrupting traditional counterfeiting workflow, Part 1: Colour and split fountains. Keesing Journal of Documents & Identity, Vol. 49, pp. 14-19.

2 Zlotnick, J., Brough, J. and Eberhardt, T. (2016). Interrupting traditional counterfeiting workflow, Part 2: Monochrome colour gamut. Keesing Journal of Documents & Identity, Vol. 51, pp. 22-27.

3 Zlotnick, J., Brough, J. and Eberhardt, T. (2017). Interrupting traditional counterfeiting workflow, Part 3: Security design and ink opacity. Keesing Journal of Documents & Identity, Vol. 53, pp. 22-28.

4 Zlotnick, J., Brough, J. and Eberhardt, T. (2017). Interrupting traditional counterfeiting workflow, Part 4: Split fountains redux. Keesing Journal of Documents & Identity, Vol. 54, pp. 31-35.

5 Zlotnick, J., Brough, J. and Eberhardt, T. (2018). Interrupting traditional counterfeiting workflow, Part 5: False split fountains. Keesing Journal of Documents & Identity, Vol. 55, pp. 26-31.

6 Zlotnick, J., Brough, J. and Eberhardt, T. (2018) Interrupting traditional counterfeiting workflow, Part 6: Split fountain position. Keesing Journal of Documents & Identity, Vol. 56, pp. 26-32.