MEASURING METALLICS

Adding a small amount of metallic lustre to a package can subtly convey the impression of a high-end or luxury product. Furthermore, the position of the metallic on a package will glint as a potential customer moves past, thus making it eye catching. Many brands believe that the extra cost of metallic ink or foil is an investment that is paid back by increased sales.

Martin Habekost of Toronto University has led research into process control for metallic inks. At three separate Technical Association of the Graphic Arts (TAGA) conferences, Habekost has compared various spectrophotometer measurement modes to the amount of ink that has been applied. There has been one significant and consistent finding – a strong correlation between metallic ink film thickness and measurements made using the M3 mode of a 45:0 spectrophotometer.

There are currently four measurement modes for 45:0 spectrophotometers in the graphic arts. They are conveniently named M0, M1, M2 and M3:

M0 Mode – First of all, the M0 mode is a catch-all for any spectrophotometer designed before the four modes became standardised.

M1 Mode – In the M1 mode, the light source in the device must match the D50 illuminant that is commonly used in light booths. Most importantly, light sources must provide the proper amount of violet and near ultraviolet light. This part of the standard was put in place to ensure that different M1 devices will agree on measurements of print on paper with fluorescent whitening agents (FWAs).

M2 Mode – The M2 mode goes to the other extreme. In the M2 mode, the light source for a spectrophotometer must have practically no UV or deep violet light. This mode will likely provide the most consistent measurements, but it discounts the effect of FWAs. Readings may not reflect what a sample looks like where the illumination is high in UV, such as outside.

M3 Mode – The M3 mode uses a pair of polarising filters to eliminate surface reflection from the measurements. An M3 mode device measures only the light that enters the ink layer and subsequently finds its way out. For glossy (very smooth) surfaces, the surface reflection is experienced as a white glint – subconsciously identified and discounted as not being part of the colour. For matte (rough) surfaces, reflection is at such a microscopic level that it is always perceived as part of the colour of the surface. This is why high densities on uncoated stock cannot be reached.

The use of polarised filters goes back to the days of densitometers. This technique was found to be useful for measuring cold-set printing on newsprint. When newsprint is ‘hot off the press’, the surface of the ink is still smooth, making it somewhat glossy. As the ink slowly dries, it conforms more to the surface of the paper fibres and thus becomes matte. The very unfortunate result of this physics is that the density of a cold-set matte print will drop significantly over the span of hours as it dries.

Researchers at Gretag, Switzerland found that measuring density with one polarised filter on the light source and one on the detector (with one rotated 90º from the other) will eliminate surface-reflected light from the measurement. With the filters in place, the density ‘hot off the press’ and the density three hours later will be the same.

Habekost’s work – and the earlier work of Jorg Mannig and Ray Verderber (see Bibliography) – have shown that elimination of this surface reflection will greatly improve the correlation between measurements and ink-film thickness. M3 mode turns a spectrophotometer into an ink-film-thickness-o-meter for metallic ink.

As an example of process control take an M3 measurement of a black patch on a matte stock. The press operator may have measured a density of 1.73D when the ink was wet 
and 1.71D later, when the ink had dried. The specification read as a satisfactory 1.75D ± 0.05.

However, in this case, the customer may not be happy if they thought that the job would be run on a glossy coated stock. Getting a density of 1.71D on a coated stock would provide a very rich black and that is the colour that was expected. But the actual print on an uncoated stock might have a density of 1.20D offering a dark grey at best. This is a failure of quality assurance.

M3 mode measurements correlate well with ink film thickness. However, the underlying issue is that the measurements do not always correlate well with the appearance of the print. This is true for regular ink, but is especially true for metallic ink. When a customer pays a premium for metallic ink, they pay the extra money for the metallic lustre. 

The M3 mode eliminates the surface reflection, yet all the information about metallic lustre is in that surface reflectance. M3 mode is an excellent metric for process control, but it is antithetical to quality assurance.

In the first image of the Lindt chocolate bar packaging, an air of elegance exudes as light reflects from the laminated foil. In the second, it is very hard to tell that there is a metallic foil. The brand name, once portraying elegance and luxury, is now a dull yellow or tan. 

The initial image was taken with the lighting and camera placed at nearly opposite angles. A tactical flashlight was used for the illumination. This can focus the beam of light so that it is virtually all going in the same direction. The camera was positioned opposite the flashlight, which created the best metallic effect. This position is called the specular angle, where a mirror image would be directed.

The second image – that would linger on the shelf for longer – was taken with the flashlight alongside the camera. A matte white sheet was placed behind the package to produce an additional diffuse illumination. Both of these reduced the amount of surface reflection seen by the camera.

The appearance of metallic lustre occurs when two things fall into place. Firstly, the surface reflection must not be scattered. If the surface is illuminated from 45º (for example) all the light that reflects from the surface must reflect at –45º. This means the surface must be smooth. Secondly, a great deal of light must reflect from the surface. The smooth surface of a windowpane reflects somewhere around 5% of light at the surface. Even though all that light reflects in the specular direction, no one would say that a window has a ‘metallic lustre’. For a surface to look metallic, the surface reflectance must be much greater such as 70%, 80% or 90%.

When it comes to correlation with human perception of metallic lustre, a gloss meter is a big improvement over the 45:0 geometry. As shown in the simplified diagram, there is a collimated light source on one side. This light source illuminates a certain area of the sample. For a perfectly flat surface, the surface-reflected light will continue to be collimated as shown in the diagram. This surface-reflected light will be captured by the detector.

If the surface is not perfectly smooth, some surface-reflected light will bounce back at different angles and will be blocked by the aperture. It will not reach the detector. Thus, the amount of light captured by the detector of the gloss meter depends on the smoothness of the surface.

The amount of light reaching the detector will also depend on how much of the incident light reflects from the surface instead of entering the sample. For printing ink, the percentage that reflects from the surface is only around 5% like glass. For a metal, far and away the majority of the incident light is reflected at the surface. Therefore, the amount of light captured by the detector of the gloss meter also depends on how well the surface is covered by the metal.

The diagram is a simplification of a drawing from ASTM standard D523 which defines three standard configurations for a gloss meter. These configurations are characterised as 20°, 60° and 85°, referring to the length of the blue arc. For glossy samples, the 20° option is recommended, with the illumination coming almost directly from above. According to the standard, the aperture will collect light through a 4.4° aperture centred at 20°.

From this description, it sounds like a 20° gloss meter might be ideal for measuring the metallic lustre of metallic ink. The gloss meter is looking at light in the proper direction to see metallic lustre. Reasonably inexpensive gloss meters are available from many vendors.

On the other hand, there are two physical effects that feed into an appearance of metallic lustre. A gloss meter combines these two physical effects into one. In the best-case scenario, the way they are combined and the size of the aperture will give a good indication of how metallic a surface appears. Nevertheless, a gloss meter might read that a polished white tile looks metallic. On the other hand, it might not be able to distinguish between the metallic-ness of two samples that are obviously different.

Research going back over 50 years – see John S Christie in Bibliography – has found that a simple gloss meter is inadequate to quantify metallic lustre. This 1969 paper describes various instruments that make measurements using multiple apertures at the specular angle and near that angle. This work led to instruments that measure quantities called ‘haze’ and ‘distinctness of image’ (DOI). These are officially defined in ASTM standard E430, specifically aimed at measurement of bright metallic surfaces. 

There are several different configurations described in this standard, but one definition of haze is based on measurements with three 1.8° wide apertures at – and at either side of – 20°, instead of a single aperture that is 4.4° wide. Haze is computed by comparing the amount of light in the centre aperture versus the ones on the sides, providing a better assessment of the metallic lustre. The collection of multiple measurements relates more directly to the two physical causes that make something look metallic.

One potential issue is that the haze meter requires a very flat sample. If the sample is warped by only 0.9°, the specular reflection will be displaced by twice that much. This moves the peak of the glint completely into the next aperture. These instruments were originally designed for measuring rigid metallic surfaces but, even still, the standard refers to a metal clamp to help flatten the surface. This is a bigger problem for measuring metallic ink. Maintaining a flat surface is considerably more difficult with normal print substrates.

The instruments described in E430 were state-of-the-art circa 1960. But today, single-point detectors can be readily replaced with imaging technology. A device with three fixed apertures can be replaced by an imaging detector that measures the amount of light at tens or hundreds of positions per degree. This allows for a ‘virtual aperture’, where the precise position of the aperture can be adjusted in software to account for the sample not being completely flat.

The curves in Figure 1 were created from measurements of a ‘sheen chart’ of Behr paint samples as measured with a Rhopoint IQ gloss/haze meter. Each curve is a plot of the amount light reflected at each angle between 12–26° at a resolution of 0.03°. 

From such data, it is not only possible to correct for the tilt of the sample, but also to emulate an instrument with any combination of apertures conceivable. Although this can be beneficial, it misses an opportunity. The researchers of the past had to be content with a small number of detectors and designed their metrics accordingly. With vastly more information, it is possible to develop metrics that quantify precisely what we seek to measure.

Figure 2 shows measurements of reflectance near the specular peak of the high gloss paint sample in green and a fit of that data to a Gaussian function, commonly called the ‘bell curve’, in blue. The fit is quite good, especially near the peak.

The parameters of this Gaussian function can be used to determine two quantities, may prove very useful for process control of metallic inks. The width of the Gaussian is the degree that the sample has disturbed the collimation of the light source, which is directly related to the smoothness of the ink. The area under the Gaussian curve represents the total amount of light that was reflected at the surface of the ink.

It was stated before that the appearance of metallic lustre occurs when two things fall into place – the surface must be metallic, so there is more total surface reflection than from, for example, non-metallic ink and the surface must be smooth so that it does not disrupt the direction of light. These properties cannot be measured with a 45:0 spectrophotometer. John Seymour believes that this analysis will provide objective measures of precisely those two quantities.

Before embarking on a career as a consultant and adjunct professor, Seymour played a leading role for 25 years in QuadTech’s advanced product development group, where he developed algorithms and instrumentation for colour management of printers. Prior to that he was a computer programmer in a variety of scientific fields. 

• ASTM, Standard Test Methods for Measurement of Gloss of High-Gloss Surfaces by Goniophotometry, ASTM E430-23
• ASTM, Standard Test Method for Specular Gloss, ASTM D523-14 (reapproved 2018)
• John S Christie, An Instrument for the Geometric Attributes of Metallic Appearance, Applied Optics, Vol. 8, Sept. 1969
• Martin Habekost and A Andino, Metallic ink measurement using the M3 mode, TAGA 2016
• Martin Habekost and Xiaoying Ma, M3 is for controlling metallics, TAGA 2017
• Martin Habekost and Xiaoying Ma, Visual and Numerical Evaluation of Metallic Inks and How They Compare to Numerical Colour Differences, TAGA 2018
• Jorg Mannig and Ray Verderber, Improving Metallic Ink Printing through Polarised Densitometry, TAGA 2002, p. 33