University of California Transportation Center

Designers are often asked to gauge the influence of an illuminant or a visual signal on the human eye in a particular application. It has long been known that physical measurement is insufficient for this purpose. This is, in part, because the human sense of vision is not equipped with a competent radiation detector that is independent of wavelength.

"The ability to see is a partnership of light and sight..."

Photometry, a specification computed from a radiometric measurement, was developed as an attempt to capture what is human about human vision so that the sterile specifications of radiometry might be thus rendered more useful as a predictive tool. As it has developed and matured, photometry has become an exacting tool requiring sometimes extraordinary attention to detail that lies beyond the experience of many of us. But even with appropriate attention to detail, photometry takes account only of the spectral distribution of radiation in comparison to the spectral sensitivity (luminous efficiency) of the observer, and the size of the pupil. These characteristics of vision arise early in the perceptual chain. Other features of human vision, which vision science is beginning to systematically quantify, may be viewed as additional signposts pointing to the need for a more thorough humanization of radiometry. Visual illusions provide an example as do a variety of visual effects in which the percept is not veridical. This paper will review selected characteristics of human vision that point to the need for moving beyond photometry. Two examples, one involving the use of LED illuminants in traffic signals, and the other showing the theoretical advantages of a modified wing-tip strobe, will be used to illustrate the problem area. It is possible to foresee a future in which designers could use a photometric correction incorporating these additional factors.