The infrared (IR) region of the electromagnetic spectrum is located beside the visible colour red, at wavelengths just longer than visible light. IR can penetrate through paint, with the depth of penetration depending on paint thickness, paint composition, and the wavelength of IR light applied. Longer wavelength IR light can penetrate through paint layers, but will be absorbed by blacks, a quality that enables researchers to use IR to look below the paint layers.
In technical examinations, two distinctly different imaging methods employ IR: infrared photography (IRP) and infrared reflectography (IRR). IRP can be done with an adapted photo camera with a charge-coupled device (CCD), but such a camera is only sensitive up to around 1000 nm. At this wavelength, lighter colours, such as reds and whites, will transmit the IR wavelengths, but to penetrate greens and blues an IRR camera is needed. The spectral response of IRR cameras varies with the elemental makeup of the specific detectors. The OSIRIS IRR camera at Queen’s University is outfitted with an InGaAs detector, which cuts off at 1700 nm.
IRR is especially well equipped to reveal underdrawings in early Netherlandish paintings, since the artists of these works would lay out their compositions in a dark, carbon-containing material on a highly reflective, whitish ground layer. The study of the style and method used to create underdrawings has been used to support or refute attributions; it can also help in mapping out the creative process, and, for example, to distinguish an original composition from a copy, or to establish patterns of collaboration within a workshop. Since darker colours are revealed, the method is complementary to X-radiography, which typically registers the lighter colours, such as lead white.
Ultraviolet (UV) light is one of the most common lighting tools by which to analyze paintings and other works of art. UV light applications can be differentiated into UV-fluorescence and UV-reflectance techniques. The fluorescence technique involves UV illumination of material, which stimulates fluorescence at a longer wavelength than the excitation source, with the resulting fluorescence typically being visible light. The cultural heritage field uses both primary and secondary fluorescence microscopy, which is based on this process. When the material itself fluoresces it is called primary fluorescence, while secondary fluorescence uses fluorochromes to tag specific non-fluorescing materials, allowing them to be identified. Reflected UV imaging involves illumination of an object with UV light, which then reflects off the object and is recorded by a UV-sensitive detector array. Interestingly, UV applications are also an extremely common means to perform forensic investigation of crime scenes.
Under UV light, pigments and varnish layers fluoresce differently, allowing for the detection of areas of retouching, since recently applied materials absorb UV and will appear relatively dark. For example, aged natural-resin coatings, such as dammar or mastic, look yellow or brown under visible light, but glow yellowish-green under UV light. Aged retouching, however, often does not show under UV. This fluorescence can be photographed with a normal camera and a special filter. Examination under UV can sometimes also provide additional information to identify certain pigments and the prior cleaning history of a painting.
Because of its multi-elemental and nondestructive capabilities, X-ray fluorescence (XRF) is a powerful and widely used technique in the field of cultural heritage. The instrument works through emission of a beam of high-energy photons that interact with subatomic components (the electrons) in the atoms in the paint, causing them to emit X-rays having energies that are characteristic of the specific atoms present. XRF gives us a nondestructive window into the elemental composition of artworks including, for example, the paints and grounds.
Qualitative XRF analysis can be used to identify pigments in a particular sample site. XRF is able to detect and identify heavier elements found in many pigments. The instrument used for this study is capable of accurately measuring the chemical elements calcium through lead. For example, the red pigment derived from cinnabar is a mixture containing both mercury and sulphur. Azurite is a blue-coloured mineral consisting mainly of copper. These are just two examples where XRF can assist in determining the pigments used. XRF can also offer semi-quantitative data about the relative amount of an element within the sample.
In the past, XRF was micro-destructive, requiring a sample to be taken from the painting and brought to the instrument. These stationary machines, which are still in use, have much smaller sample sizes than the more recently developed handheld XRF analyzers, and are useful when more precise analysis is required.
A handheld XRF device has the double advantage of being both noninvasive and able to analyze a painting in situ. This makes it an attractive method for conducting an initial pigment analysis; however, the new portable XRF machines are not without their drawbacks. Many instruments have a relatively large sampling area (anywhere between 5–10 mm), and as such the resultant XRF spectrum may contain a number of pigment mixtures, making it difficult to interpret the elemental data for any one pigment. Another complicating factor is the penetration capability of the X-ray beam. Depending on this penetration potential and the sample composition, more than the topmost layer of the painting may be sampled, resulting in a spectrum that includes data from layers below the surface. Since paints are often highly complex mixtures, the interpretation of the spectra requires the involvement of a specialist.
Co-registration of Different Imaging Methods
Image registration is the alignment of two or more digital images of the same object. Multiple images can be captured repeatedly using the same method, such as using a digital camera to take pictures that are later processed for image stabilization; or the images can be taken using different imaging devices or sensors. Multiple digital images of an object contain complementary information such as edges in the picture or image brightness/darkness. This similar information content is used to find the best mathematical way to align the images. To perform co-registration of images of the paintings taken with different imaging methods, custom in-house software was written using the commonly used open-source Insight Segmentation and Registration Toolkit. Image co-registration used mutual information as a similarity metric and an affine transformation was used to register each type of image. The procedure was difficult because of the immense size of the computer image files, and image contrast from the various methods, particularly from X-ray images, made choice of a similarity metric a challenge.