Date of Award


Degree Name

Doctor of Philosophy




Carl W. Dirk


The purpose of this research is to design the best lighting that will minimize long term photochemical degradation of Old Master Drawings and/or other works of art, while maintaining the patron's appreciation of the object's color and detail. The present approach is a technological refinement to the basic underlying earlier work on fluorescent lighting by W. A. Thornton, W.A. at General Electric1. Thin-film dielectric, multi-coating technology is used to create filters that eliminate ultraviolet light, near infrared light and significant unnecessary parts from the visible spectrum, thus maximizing the reduction in photochemical degradation, while maintaining optimal color rendering. Three interference filters, were designed and manufactured successfully. The filters are designated Mark 1, Mark 2, & Mark 3. In this dissertation, the filters are analyzed with regard to their performance parameters. This includes color rendering, retardation in fading or color change, beam angle effects, filter stability, perceived brightness, and visual appreciation parameters.

To a high confidence level, all three filters are perceived as being indistinguishable from unfiltered light with regard to the color confusion index parameter (CCI). Subjective assessments by tests subjects suggest the Mark 3 filter may display some distinguishability with a confidence level for distinguishability of 35% for the overall satisfaction parameter. The Mark 3 filter is the most complex three-color type spectral profile and this might be expected due to beam angle effects or departures in accuracy of color theory. Beam angle affects suggest that the Mark 1 and Mark 2 filters do not display significant color rendering aberrations due to Newton's colors interference effects, except possibly at the periphery of the broadest (55-60+°) beam angle lamps.

Filtered and unfiltered light are effectively of the same perceived brightness, though to a low confidence level, unfiltered light might be perceived brighter. Accelerated aging studies of the filters indicate useful mean time before failures of >20 years. In fact, no failure was observed in any of the accelerated studies, The Mark 2 and Mark 3 filters were evaluated in equal-luminosity studies with regard to their effect on limiting fading of both standard fading samples such as the ISO Blue Wool series, and also other commercial pigments or stains. Within experimental error, the Mark 2 filter either slowed fading or had no effect on fading for all pigments relative to unfiltered light and Optivex® filtered light, Optivex® being a common commercial filter used to protect works of art. Mark 3 protected in many cases, but for some pigments it was less protective than Optivex® filtered light. This failure is interpreted in terms of the wavelength dependence of the excess light in certain wavelength regions on an equal luminosity condition, and suggests that more subtle wavelength dependent optimizations have to be undertaken for filters which possess significant band separation.




Received from ProQuest

File Size

429 pages

File Format


Rights Holder

Monica Fabiola Delgado