Search

A series of experiments is described, the aim of which is to identify the optimum spectral band and detector configuration for the visualization of underdrawings. First, the visualization of underdrawings is investigated using three detector systems operating in different infrared spectral bands from 0.9 to 2-5µm. Visibility analysis is applied to the examination of eight test panels, each having a rectangle of paint covering six different underdrawing materials. These data demonstrate that the combined optical properties of the pigment and the underdrawing material determine which spectral band gives the optimum visualization of the underdrawing. The peak visibilities of different combinations occur between 1.0 and 2.5gm. These results suggest that camera systems having uniform spectral responsivity across the l.0 to 2.5pro band would be most useful for infrared reflectography (IRR). Second, the imaging properties of each camera system are investigated. Comparison of the modulation transfer function (MTF) and signal-to-noise ratio for the vidicon and the platinum silicide (PtSi) camera systems shows that the latter is much closer to a theoretically perfect system. Finally, the applicability of the visibility studies and MTF modeling to the IRR application is demonstrated by the examination of two paintings with the PtSi camera, which renders dramatically improved images of underdrawings and paint changes, as compared to reflectograms obtained with the vidicon camera. The implications of this study, including the potential use of infrared reflectography as a method of non-destructive analysis, are discussed. [ABSTRACT FROM AUTHOR]

A method is described for the evaluation of the performance of near-infrared imaging devices to detect underdrawing and reworking of paintings. System analyses have been performed on two different near-infrared imaging devices: a Hamamatsu lead sulphide vidicon camera and a Photometrics silicon charge coupled device (CCD) camera. The analyses included modelling of the system spectral response curve by analysis of the system components, as well as considering the limiting resolution and modulation transfer function (MTF). The modelling was confirmed by examination of a test panel at various wavelengths. Calculation of visibility is introduced as a quantitative measure of system performance. This calculation is also discussed as a means to define more clearly the actual performance requirements of these systems. This analysis has revealed the strengths and weaknesses of current systems, and provides guidance in the construction of improved devices. [ABSTRACT FROM AUTHOR]