An ongoing concern of conservators when restoring paintings and polychrome objects is the possibility of irreversible
damage to the original paint layers. This is the case when removal and replacement of aged varnishes is being
considered. Although much work has been conducted to determine the effect of the removal of a varnish on the condition
of the (painted) surface, there is still some uncertainty in the conclusions because it has, until recently, been impossible
to characterize the original surface under the varnish in situ. A promising solution to this problem is the use of white
light confocal profilometry. An initial study has been conducted using this technique to measure the roughness of pencil
lines and oil paint under a varnish. Measurements were made on areas up to several mm<sup>2</sup> in size, at lateral resolutions
down to 1 μm, and vertical (roughness) resolutions of 0.1-0.3 μm. It was found that the surface roughness of surfaces
under relatively thick varnishes can be measured at relatively low magnification (20x). For thin varnish layers typical of
paintings, higher magnifications are required. White light confocal profilometry has thus been shown to be a useful tool
for <i>in situ</i> studies of the (sub)surface properties of (painted) objects.
For archiving, art historical purposes and restoration, "two dimensional" (2-D) works of art such as paintings or photo-graphs have, in modern times, traditionally been documented using (colour) photography and/or digital imaging tech-niques. While current technology allows reproductions with high spatial and colour resolution, these techniques only document the 2-D form and colour of the surface of the object. Surfaces also have, however, three dimensional (3-D) properties which play just as important a role in determining an objects appearance and how it is perceived. Among these, surface roughness/texture and the optical properties of transparent layers are particularly important. Both determine the way light is scattered from the surface, and influence not only colour perception, but also glossiness, illusions of transparency, and depth perception. These properties are often the first to be affected before colour changes happen or are perceived in a measurable quantity. They are difficult, if not, impossible to document using current 2-D techniques, while current 3-D scanners do not have the resolution necessary to document the micro-roughness of surfaces which actually determine appearance. The Netherlands Institute of Cultural Heritage has started a multidisciplinary programme to study the effect of 3-D pro-perties of surfaces on the appearance of works of art. The combined effects of colour, roughness, and optical properties of transparent layers are being investigated in situ using standard engineering micro-roughness measurements, colour spectroscopy, and digital imaging techniques. Initial work indicates that these are excellent methods for documenting, for example, the effects of cleaning of paintings and face-mounted photographs. Combined with light scattering models, they will provide a good tool for understanding the appearance of objects, and for their conservation. Rendering models, for example, could help conservators make selections of retouching materials based on colour and texture, or varnishes based on optical properties.