Katie Sanderson, Assistant Conservator of Photographs at the Metropolitan Museum of Art (MMA), presented a most informative comparative study of handheld spectrophotometers undertaken at MMA. When the Department of Photograph Conservation decided to replace its existing handheld spectrophotometer—an X-Rite 968—Sanderson along with Scott Geffert, Senior Imaging Systems Manager, researched current units available to determine the best replacement and variation in measurements taken by each.
Sanderson began by outlining the factors to consider when replacing a spectrophotometer: data continuity (extant data over 20 years); instrument agreement; data translation; software compatibility with previous and future instruments; and longevity and support (the previous spectrophotometer is no longer supported by X-Rite but still takes good data readings). In total, seven spectrophotometers—four by X-Rite and three by Konica Minolta—were tested against the MMA’s X-Rite 968 and a bench-top spectrophotometer equipped with an external remote diffuse reflectance accessory probe in the Department of Scientific Research. The seven spectrophotometers examined were:
- X-Rite 964
- X-Rite eXact
- X-Rite Ci64
- X-Rite RM200
- Konica Minolta 2600D
- Konica Minolta 2500c
- Konica Minolta FD-7
Before delving into the specific finding of each unit tested, Sanderson provided a brief overview of how spectrophotometers work. She explained that an object is illuminated by a light source of a specific spectral range, a detector collects any reflected light, and a unique spectrum is produced. While some light sources extend into the ultraviolet region of the electromagnetic spectrum, most are within the range of visible light (400-700nm). The two most common geometries for spectrophotometers are 0/45—in which the first number represents the angle (in degrees) of the light source and the second number the angle of the detector—and integrated spherical.
As Sanderson described, some of the units tested had an integrated spherical geometry that takes into account specular reflectance; these spectrophotometers can be operated in either specular component excluded (SCE) or specular component included (SCI) mode. The aperture of the X-Rite units was set to 4mm—that of MMA’s current spectrophotometer—andthe Konica Minolta units were set to 8mm as they exhibited a range of apertures and, in some cases, were not adjustable.
To determine an appropriate replacement, several reference standards and sample objects were tested with the seven spectrophotometers. The reference standards tested were obtained directly from X-Rite that is about to release a new Digital SG ColorChecker. The new target will include the same colors as the existing one but will utilize new pigments for some colors. For this comparative study, MMA obtained samples of the new standards to assemble its own large-format color checker. Ceramic BCRA calibrationcolor tiles were also tested as well as objects with varied surface qualities—chromogenic photographic prints (glossy and matte), watercolorpaper, textiles, and paintings. Five readings were taken and averaged for each spot tested; the units were lifted and repositioned before each measurement to account for a margin of error in positioning when monitoring color shift in objects over time using a spectrophotometer. Mylar® templates were created to facilitate positioning of the meters. All testing was completed by a single operator and resulted in approximately 12,000 readings!
To evaluate the variation in measurement between spectrophotometers, MMA’s X-Rite 968 was used as a master and delta E values were calculated for each of the 140 X-Rite color references. Sanderson summarized the results of this comparative study as follows. Meters with a 0/45 geometry produced readings with the closest match to the unit currently in use, which was not surprising as both are 0/45 instruments. When operated in SCE mode to exclude specular reflectance, the integrated spherical instruments fared worse than the 0/45. The easiest-to-use instruments were lightweight with built-in crosshair targets to facilitate alignment with a template. Finally, Sanderson introduced the concept of acceptable tolerance meaning that the operator should simplify the use of spectrophotometric readings by using a single instrument with a single set of standards. During the Q&A session that followed this presentation, a member of the audience asked which spectrophotometer MMA ultimately selected. Sanderson responded that the X-Rite eXact was selected for several reasons: it is lightweight; it produces data reasonably consistent with MMA’s existing spectrophotometer (understanding that data translation will be necessary regardless of which instrument is chosen); and long-term support from the manufacturer as well as continuity in data and software.
The presentation concluded with a discussion of areas of further research within this project, specifically continued analysis of data pertaining to the UV-radiation source found in some of the meters as well as the use of SCE settings in spherical integrated systems for more highly textured surfaces like those found in textile objects. Finally, it is a goal of MMA to complete processing of all data collected during this study and make it available to a wider audience so that it might contribute to more standardized color communication within the field of conservation and allied professions.