Mesopic Vision - Road Lighting: Fundamentals, Technology and Application - page 79

The human crystalline lens in the eye turns yellowish with age, which means that
blue light is largely absorbed in the lens (see Chap. 7 “Age effects”). This implies
that the mesopic advantages of blue-rich light, as described above, are smaller for
older drivers.
The method for determining the actual adaptation luminance in a non-uniform
environment has not yet been defined and standardized. As has been discussed in
Sect. 6.4 it is recommended that at least 25 % of the average road-surface luminance
be added to the actual average road-surface luminance when estimating the adaptation
luminance.

6.5.2 Brightness

6.5.2.1 Models for Brightness Prediction

It has been mentioned in Chap. 5 that quite a few investigations have shown that
the spectrum of a light source also influences the subjective impression of bright-
ness (Morante 2008 ; Rea et al. 2009 ; Knight 2010 ; Schäfer et al 2010 ; Fotios and
Cheal 2011 ). This is not a specific property of mesopic vision, but holds true for
both mesopic and photopic vision. In the mesopic vision range, light sources with
a relatively large proportion of light emission at short wavelengths viz. with high
colour temperatures and S/P ratios, produce higher brightnesses for the same illu-
minance. CIE ( 2011 ) developed a “supplementary photometric system” with which
light sources or lighted objects can be evaluated in terms of comparative brightness 6 .
Other models based on the possible interactions of the different types of light sensi-
tive cells are being discussed (Fotios and Cheal 2011 ; Rea et al. 2011 ; Bullough et al.
2014 ). Fotios and Cheal ( 2011 ) tested the brightness prediction of different models
against brightness assessments made for those lamps, typically used in road lighting.
The lamps were compared in two side-by-side booths. The reference luminance was
0.25 cd/m 2 . Not surprisingly, the CIE supplementary system for brightness evaluation
showed a good prediction (correlation coefficient R 2 0.89). Interestingly enough,

also the CIE system described in the previous section, developed for mesopic task
performance, was also tested. It showed a good prediction with a only slightly lower
correlation coefficient of R 2

=

= 0.86.

6.5.2.2 Consequences for Road Lighting Practice

For practical reasons of having to use only one system for road lighting, the proposal is
that the same CIE system be used for both peripheral task performance and brightness

6 The photopic and scotopic luminance together with the chromaticity coordinates x and y (as also
used to position lights in the colour triangle) are the quantities used.