Abstract

Painting conservators are often tasked with creating precise color matches. Though conservators often use natural daylight, natural lighting can be unreliable. Conservators may turn to additional light sources to balance weather conditions, to extend working hours, or to increase lighting intensity.  This research considers the color temperature of light sources used during color matching and evaluates their impact on the conservator’s ability to create in-painting which remains consistent and seamless when viewed under a wide range of light sources with different color temperatures.

​

Experimental Design

All light sources emit visible light, but we do not see their light as the same color. Radiation emitted from each source varies across the visible spectrum and results in a different spectral power distribution and perceived color temperature. Our perception of color results from the interaction of a light source, an object, and an observing system. Any change in these three factors can cause shifts in the perceived color. My research aimed to control the objects and viewer, while introducing variation within the light source.

 

Blue is often a problematic color during in-painting. For this reason, blue pigments and paints were selected for this experiment as they are good indicators of color shifting caused by variation in color temperature. Paint samples with artificial losses were made in three testing groups: lapis pigments, azurite pigments, and mixed or manufactured paints. The duplicate losses in each paint swatch were in-painted under the illumination of light sources with a notable difference in color temperature. In-painting was performed under 3000K and 5000K illuminants on half of the samples. The second half were matched using illuminants at 4000K and 6000K, in natural daylight.  Finally, the in-painting completed on the samples was visually compared and photographed in each of the tested color temperatures to evaluate the consistency of the match.

​

Conclusions

Based on my  observations, when not supplemented daylight illumination can lead to problematic matches for blue pigments. The natural variability of daylight rendered surfaces under a wide range of measured color temperatures. However, this variability can also lead to pitfalls. On a cloudy day, daylight conditions measured 1000K cooler than the normal average 6000K, during sunny spells, and up to 5000K cooler when cloudy. In these fluctuating conditions, additional lighting is beneficial for both variety in color temperature and increased illumination intensity.

​

The color temperature of a single illuminant was shown to have significant impact on both the process of in-painting and the constancy of  the achieved color matches. Among the tested illuminants, the 4000K LED and natural daylight generated matches with the best color constancy. When used together, these sources provided a well-balanced lighting environment. For specific instances where daylight causes known color shifting, it may be preferable to complete in-painting with the addition of a warmer color temperature source to avoid poor color matching.

​