John Seymour

Quantification of Metamerism In the Graphic Arts

John Seymour, John the Math Guy, LLC

Metamerism refers to the fact that two reflective samples may match under one illuminant, but not under another. Metamerism is due to the fact that human vision reduces spectra to three dimensions; one dimension for each type of cone in the eye. This loss of information means that a variety of different spectral curves may translate to exactly the same visual response.

The use of expanded gamut printing to replace spot colors has demonstrated huge cost savings for the printer. But metamerism is one consideration that may be neglected. Despite excellent color management, if a package printed with a spot color (which was mixed in the ink kitchen) sits on the shelf next to another package printed with a combination of CMYKOGV inks, the potential exists for the packages to match under one lighting, but not under another. Is this source of metamerism large enough for concern?

Another under-appreciated form of metamerism shows up because viewing booths are necessarily a simulation of the standard D50 illuminant. On the other hand, software that translates spectra into CIELAB values will use the theoretical D50 curves from CIE 15. When excellent color management is practiced, a colorimetric match may be found according to the numbers computed with official D50 standard, but there may not be a visual match under simulated D50. Is this source of metamerism large enough for concern?

Several sets of metamers have been created by the CIE to assess how well one illuminant matches a standard illuminant [CIE 1995, 1999, 1999, 2004]. These sets are small collections of pairs of spectra which match perfectly under certain standard illuminants such as D65 or D50. A metric known as the metameric index is used to quantify the quality of the simulation of a standard illuminant. It is computed by computing the CIELAB values of the metamers from the collection under both the standard illuminant and the test illuminant. The metameric index is a combination of the color differences due to illuminant.

The utility of the metameric index is based on the assumption that if the test illuminant performs well on the small set of metamers, it will also do well out in the field for different metamers. This assumption depends on the degree to which the spectra in the metamer set resemble the spectra of the real-world application. Therefore, in order to assess metameric suitability, it would be useful to create a database of metamers that is based on the spectra of actual print.

In the first part of this project, a set of metameric pairs under D50 was created. Measurements from a Pantone book were used to define target L*a*b* values.

The 2019 Ryerson University Expanded Gamut Study provided data that was used to derive metameric pairs for the Pantone spectra. Profiling charts were printed on an HP Indigo 7900 press. For those Pantone colors in the CMYK gamut, spectra of CMYK overprint combinations were found which came close to the target L*a*b* value. These CMYK spectra were digitally tweaked to make a perfect metameric match. This was repeated for expanded gamut (CMYKOGV) data.

This metamer database can be used to answer practical questions about metamerism. In this paper, the metamer database is first used to assess the degree of metamerism that can be expected when expanded gamut is used to replace spot colors. The metamer database is then used to determine the degree of metamerism that is caused the simulation of D50 caused by real viewing booths.

Bibliography

CIE (1995), Method of Measuring and Specifying Colour Rendering Properties of Light Sources, Publication 13.3

CIE (1999), A Method for Assessing the Quality of Daylight Simulators For Colorimetry, Publication 51.2-1999

CIE (1999), Colour rendering (TC 1-33 closing remarks), Publication 135/2

CIE (2004), CIE Colorimetric and Colour Rendering Tables, Disk D002, Rel 1.3

Sharma, Abhay, Ryerson University Expanded Gamut Study 2019, Evaluation of Spot Color Reproduction in Multicolor Printing, Version 2.7, 9 September, 2019