Automatic white-balancing algorithms play a key role in digital photography. Failure to estimate illumination chromaticity correctly will result in an invalid overall color cast in the final image. Such image-wide color casts are easily detected by human observers. The aim of this paper is to quantify the subjective effects of white-balancing errors in digital photography. It is achieved by means of subjective tests. Different natural images are utilized in order to study the effect of image contents on the acceptability of white point shifts. Multiple directions of white point error are covered in order to study the relation between image contents and the most acceptable direction of white point shift. Details of the camera system and sensor used are also discussed in order to separate the contribution of white balancing errors from the contribution of other color processing to the final color errors. Typically the performance of automatic white-balancing algorithms is presented in the literature by using different color or color difference measures. This paper provides a means to interpret those results from the point of view of digital photography, namely, to separate significant white-balancing errors from insignificant ones.
"Subjective effects of white-balancing errors in digital photography," Optical Engineering 47(11), 113201 (1 November 2008). https://doi.org/10.1117/1.3013232