TY  - JOUR
PY  - 2009//
TI  - New light sources, automotive lighting, and the Purkinje shift
JO  - Eye and the auto, the
A1  - Flannagan, Michael J.
SP  - 23
EP  - 23
VL  - 2009
IS  - 
N2  - CONTEXT: Prior to the mid 1980s, virtually all vehicle lighting used incandescent sources. The spectral power distribution of incandescent sources can be changed in only minor ways, and filtering is therefore the only practical option for modifying the color of incandescent lamps. In recent years, new light sources have been adopted in automotive lamps that offer more options for spectral output. High-intensity discharge (HID) sources have been used in headlamps, and light-emitting diodes (LED) have been used in signal lamps and headlamps. Because of the more flexible spectral options, it has become more important to understand how the spectrum of light from automotive lamps affects drivers‘ visual performance. In particular, it has become important to understand the roles of rod and cone receptors in night driving, and how the differing spectral responses of those receptors may influence the overall effectiveness of automotive lighting, including both exterior and interior lighting.    OBJECTIVES: The objective is to review a variety of empirical results and describe the implications of those results for how the light from automotive lamps should be measured. In particular, the results are relevant to the implications of the difference in spectral sensitivity between rod and cone receptors (the Purkinje shift) for automotive lighting.    METHODS: This presentation will review several experiments that have measured drivers‘ nighttime visual performance on the road and on test tracks. The primary independent variables are the intensity and spectral power distribution of the various light sources involved. The dependent measures of interest include seeing distance to pedestrians, masking of stimuli by sources of glare, conspicuity of signal lamps, and ratings of discomfort from glare. Analysis includes photopic and scotopic photometric measures of the light stimuli.    RESULTS: Seeing distance to pedestrians is related most strongly to photopic measures of light. Scotopic measures do not predict seeing distance, but are predictive of drivers‘ ratings of several subjective phenomena. The effects of spectral power distribution on discomfort glare are not fully accounted for by either photopic or scotopic photometry.    CONCLUSIONS: Photopic photometry is a good predictor of the effect of light from automotive lamps on the objective visual performance of drivers. This suggests that cone photoreceptors play the dominant role in this performance. Traditional (photopic) photometry is therefore appropriate for specifications relevant to headlamps. For conspicuity of signal lamps, other measures may be appropriate and should be further investigated.<p />
LA  - 
SN  - 
UR  - http://dx.doi.org/
ID  - ref1
ER  -