Surface Analysis: Introduction
Surface analysis, or the study of the outermost layer of an artwork, is a key component in the material characterization of photographs. Using a combination of human observation and newly developed automated digital capture techniques—enhanced photomicrography and polynomial texture mapping—conservation scientists at The Museum of Modern Art gathered data on the surface topography, surface sheen, and surface texture of the photographs in the Thomas Walther Collection. Much of this data, which deeply informs the research and analysis throughout the Object:Photo website, is presented on the individual artwork pages in the Photograph Index; it is also incorporated into the site’s powerful visualization tools. This section describes the surface-analysis techniques used in the project, including detailed procedural protocols.
Human-Observation Survey for Surface Sheen
Photographic paper is manufactured in a wide range of surface sheens, or gloss levels. Between 1909 and 1949, as today, papers were available with surfaces ranging from glossy (highly reflective) to matte (light absorptive), with many varieties in between. Each surface type bestows a specific quality on the final print: glossy surfaces transmit fine detail and thus were favored by press photographers looking for reproduction-quality images; matte surfaces—softer, more like drawing paper—were sought by those, like portrait photographers, aiming for a more traditional artistic effect.
For this study, Lee Ann Daffner, Andrew W. Mellon Foundation Conservator of Photographs, and Hanako Murata, Assistant Conservator of Photographs, The Museum of Modern Art, conducted a simplified human-observation survey, with the goal of identifying and categorizing the surface sheen of the prints in the Thomas Walther Collection. This classification is useful when working offsite, away from the collection, and it may inform the interpretation of the results of other methods of analysis, such as X-ray fluorescence analysis (XRF).
Three surface sheen categories were developed in advance of the survey, informed by historical marketing material published by manufacturers of photographic papers and by examination of the prints themselves:
Matte: These papers contain specific admixtures, such as silica particles, that modify the surface. They may appear to be smooth or have a subtle, pebbly texture, but they have the distinct quality of absorbing light (like velvet) rather than reflecting it. These papers are often described in the manufacturing literature as “chamois,” “suede,” or “carbon.”
Semireflective: The gelatin emulsion on these papers has a smooth and reflective surface, often with a slightly grainy texture. The surface reflects light to a moderate degree. This group has the greatest variety of levels of reflectance. These papers are often described in the manufacturing literature as “luster,” “velvet,” or “silk.”
Glossy: These papers are specially formulated to achieve a highly reflective glasslike surface with a very smooth texture, which maximizes reflectance. These papers are often described in the manufacturing literature as “glossy,” “enamel,” or “brilliant.”
Two additional categories were created during the course of the survey to accommodate the extraordinary range of photographic paper surfaces observed:
Semireflective towards Matte: These papers are not completely matte yet are distinctly more matte then the typical semireflective surface.
Semireflective towards Glossy: Papers in this category are quite reflective but do not have the glasslike texture of truly glossy prints.
The survey was carried out in June 2011, using only the human eye and natural raking light from south-facing windows. Light-filtering blinds were drawn at all times. The prints in the Walther Collection were each compared side-by-side with six prints from MoMA’s photography conservation study collection, chosen to represent the three categories; based on this comparison, each print in the Walther Collection was assigned a surface sheen category.
Polynomial Texture Mapping for Surface Topography
Polynomial texture mapping (PTM), also known as reflectance transformation imaging (RTI), is a computational photographic technique that creates interactive digital images in which the details of an object’s surface texture are enhanced. A polynomial texture map (PTM) is made from a set of around forty digital photographs of the object, each taken by a stationary camera under a different lighting direction. RTI software assembles the photographs into a fluid interactive visualization of the object’s surface in which changes in lighting direction and wavelength may be simulated.
PTM was developed to study marks on ancient stone tools found in archaeological digs, and its primary value lies in making subtle surface topographical features visible to the naked eye. The software provides a number of effects, from simple raking light to extreme manipulations, and an endless array of fine adjustments are possible. Still images can be captured at any point in a visualization.
In a PTM of a photographic print, unnoticed aspects of the object’s condition, such as subtle creases, folds, and tears, may become obvious. In other cases, features inherent to the print, such as the paper’s overall surface texture, may be amplified. In the software’s specular enhancement mode, surfaces often appear glossy and wet or have a metallic appearance, an enhancement that exaggerates very subtle surface features. Also often observed in PTMs of photographs is a sparkly (metallic or crystalline) appearance in the image silver. Perhaps most curious are the circular wipe marks that appear in PTMs of photographs that have been ferrotyped, probably a result of the transference of residue from a ferrotyping plate that had been wiped clean. These and many other observations await full-scale investigation. PTM is a new tool, and there is much to learn about it the role it can play in the surface analysis of photographs. For more information about PTM, see Cultural Heritage Imaging, "Reflectance Transformation Imaging (RTI)".
Polynomial texture maps (PTMs) were made for the prints in the Thomas Walther Collection by Lee Ann Daffner, Andrew W. Mellon Foundation Conservator of Photographs, and Hanako Murata, Assistant Conservator of Photographs, The Museum of Modern Art, with assistance from Paul Messier, independent conservator of photographs and Thomas Walther Collection conservation scholar, and Sarah Yeh, intern in the Department of Imaging and Visual Resources. Between 2010 and 2013, the works were photographed in a custom-made domed light array with forty-two affixed lights, each of which lit the work from a different angle. The shutter release on the 35mm camera attached to the array was synched with the lights, and a picture was taken of the work under each of the forty-two lighting conditions. In 2014 the photographs were assembled into PTMs with open source RTI software developed by HP Labs (see HP Labs, "Polynomial Texture Mapping [PTM]").
Texture Imaging for Surface Texture
The surface texture of a photographic paper is a decisive factor in the appearance of the final print, and papers are produced in a wide range of textures—from visibly rough to very smooth. By the early 1920s, surface texture was one variable called out in manufacturers’ marketing material. For the photographer, smooth surfaces conveyed detail clearly, while course or rough textures yielded softer images and were useful for generalizing detail.
For this study, the surface textures of prints in the Thomas Walther Collection were documented in a two-step process by Paul Messier, independent conservator of photographs and Thomas Walther Collection conservation scholar, in October 2013. First, raking light photomicrography was used to capture a sample area of the surface of each print. Then the resulting images were digitally processed to enhance the visibility of the surface texture to the naked eye.
The images were acquired with a microscope system assembled using an Infinity 2-3 imager, manufactured by Lumenera, fitted with an Edmund Optics VZM 200i lens. The imager uses an Interline Sony ICX262 3.3-megapixel color progressive scan CCD sensor, producing images that are 1,536 by 2,080 pixels, with each pixel 3.45µm square. A fixed-point illumination source with a 3 inch (7.6 cm) LED line light, manufactured by Advanced Illumination, was placed at a 25-degree angle to the surface of the photographic paper.
Each print was placed in a large protective mat-board folder, with a Mylar overlay. When the folder was closed, previously prepared holes in the Mylar and in the mat board exposed a 1.00 by 1.35 centimeter (3/8 by ½ inch) area of the print for imaging. The ideal sample site was a highlight area, such as a margin, with no image design. If such a location were not available, the operator selected an area with minimal image and a flat, even surface. The folder containing the print was placed under the microscope and lighting system, which was cradled by a four-sided black-box barrier to control the effect of ambient light. Fine focus adjustments were made and the picture was taken. This image-capture technique is noncontact and nondestructive and therefore easily adapted for use on photographic prints of high intrinsic value. It is also relatively quick and requires minimal specialized handling, so that large image sets can be produced rapidly.
Each raking light photomicrograph generated a 16-bit TIFF. To improve human discernment between the raking-light images, each image was cropped to 1,024 by 1,024 pixels and processed to remove color and equalize the histogram.