Comparisons of Scotopic/Photopic Ratios Using 2- and 10-Degree Spectral Sensitivity Curves

applied sciences

Article Comparisons of Scotopic/Photopic Ratios Using 2- and 10-Degree Spectral Sensitivity Curves

Jianhua Ding 1, Qi Yao 1,2,* and Lei Jiang 3

1 College of Architecture and Urban Planning, Shenzhen University, Shenzhen 518060, China; [email protected] 2 Beijing Advanced Innovation Centre for Future Urban Design, Beijing University of Civil Engineering and Architecture, Beijing 102627, China 3 No. 9, 518 Lane, Ji’nian East Rd, Baoshan, Shanghai 200439, China; [email protected] * Correspondence: [email protected]; Tel.: +86-1352-875-6877

Received: 12 September 2019; Accepted: 15 October 2019; Published: 22 October 2019

Abstract: Despite the fact that a 2-degree spectral sensitivity curve (SSC) is extensively used in scientific research and relevant applications, the choice between the 10-degree or the 2-degree photopic SSCs in practical applications for the calculation of scotopic/photopic ratios (S/P ratios) depends on actual needs. We examined S/P ratios for more than 300 light sources for correlated colour temperatures (CCTs) from 2000 K to 8000 K and blackbody radiant spectra from 10000 K to 45000 K using 2- and 10-degree SSCs. Results showed that the ratio of the S/P values calculated using the 10-degree and 2-degree SSCs was approximately equal to 0.916. The average mesopic luminance difference increased from 0% to 5.7% at a photopic adaptation luminance from 0.005 to 5 cd/m2. For most practical applications, the mesopic luminance values calculated using these two SSCs were different by several percentage units, yet these differences could be neglected. At extremely high CCTs over 10000 K, the mesopic luminance difference may approximate the maximum value of 16%. This work proposes the conversion coefficients for S/P ratios and the transforming mesopic luminance values calculated for 2- and 10-degree SSC systems. These results may help researchers clarify differences between the S/P ratios calculated using different SSCs.

Keywords: mesopic vision; luminous radiation eﬃcacy; scotopic vision; mesopic luminance; photopic adaptation luminance

1. Introduction There are three vision states for human beings, photopic vision with luminance values >5 cd/m2, mesopic vision with luminance values from 0.005 to 5 cd/m2, and scotopic vision with luminance values <0.005 cd/m2 [1]. The luminance of values of several scenes, like roads, tunnels, and dashboard lightings, likely fall in the mesopic vision range. To measure light source performances in mesopic vision states, the scotopic/photopic ratio (S/P ratio) is used [2]. Effectively, this is the ratio of the luminous efficacy of radiation (LER) in scotopic and photopic vision. There are two photopic spectral sensitivity curves (SSCs): one is the Commission Internationale de lÉclairage (CIE) 1924 2-degree standard observer SSC [3] and the other is the CIE 1964 10-degree standard observer SSC [4]. Some studies claim that the 10-degree SSC constitutes a better representation [5]. Prior research studies on colour differences also show that 10-degree colour matching functions yield smaller colour differences than 2-degree colour matching functions [6]. Another study [7] suggested that there may be problems in the short-wavelength part of the 2-degree SSC. In road lighting, visual tasks sometimes require large visual fields, while peripheral vision is also important [8–10]. The rod photoreceptors are mainly distributed in the part of the retina associated with peripheral vision. We may thus conclude that both the 10- and 2-degree systems should be considered for uses in various applications [11].

Appl. Sci. 2019, 9, 4471; doi:10.3390/app9204471 www.mdpi.com/journal/applsci Appl. Sci. 2019, 9, 4471 2 of 10

CIE recommended mesopic vision models for road lighting [2] and published guidelines on how to use the model [12]. However, CIE did not recommend which SSC should be adopted for calculation, implying that either SSC type can be used in applications. In most of the published research studies and practical applications, the 2-degree SSC system has been used. Berman [13] reported S/P ratios of 16 typical light sources in the range from 0.23 to 2.47. Lita et al. [14] reported a light source with an S/P ratio of 2.56 at a correlated colour temperature (CCT) of 5368 K. Nizamoglu et al. [15] reported a white-light emitting diode (WLED) with an S/P ratio of 3.04 at the extremely high CCT of 45000 K. Guo et al. [16] proposed light-emitting diodes (LEDs) for CCTs from 2700 K to 45000 K with S/P ratios in the range from 1.83 to 3.84. Zan et al. [17] reported WLEDs that reached S/P ratios from 1.54 to 3.84, which corresponded to CCTs in the range from 2700 K to 7580 K. Li et al. [18] reported two-hump quantum-dot-based LEDs that reached an S/P ratio value of 2.09 at 4038 K, and a four-hump LED that reached an S/P ratio value of 3.08 at 5015 K. Zhang et al. [19] optimised hybrid white LEDs for high S/P ratios at varying CCTs that reached high colour rendition, and compared their differences. Despite the fact that in some studies it has not been mentioned which SSC was used, we rechecked the spectra in the literature and have confirmed that most studies have used 2-degree SSCs. To date, no study has compared the differences of the calculations of S/P ratios using these two SSCs. In practical applications, there are urgent needs for large-scale assessments of urban lighting [20] and nightscape lighting. However, there are some unsolved technical issues to assure photopic adaptation luminance [21,22], which may cause inaccuracy in assessments. In this study, we examined the differences of the calculated S/P ratios with two SSCs based on tests with more than 300 light sources [23] with CCTs from 2000 K to 8000 K, and with blackbody radiant (BR) light sources with CCTs from 10000 K to 45000 K. We calculated the mesopic luminance levels at photopic adaptation luminance levels from 0.005 to 5 cd/m2 based on the use of the two tested SSCs and compared the differences. Conversion coefficients between the S/P ratio and mesopic luminance calculated using 2- and 10-degree SSCs were proposed. The work may help elucidate the S/P ratios and mesopic luminance differences calculated with these two different SSCs in research studies, and provide convenient conversions in practical applications.

2. S/P Ratio of Monochromatic Spectra

The S/P ratio was calculated by dividing the LER in scotopic vision (LERs) by that in photopic vision (LERp). LERp was defined by Equation (1), LERs was defined by Equation (2), and the S/P ratio, which can also be expressed as a quotient in the form of Rsp [12], was defined with the use of Equation (3). Herein, P(λ) represents the spectral power distribution (SPD) of the light source, Km is the maximum photopic luminous efficacy of 683 lm/W at 555 nm, and Km is the maximum scotopic luminous efficacy of 1700 lm/W at 507 nm. Accordingly, V(λ) is the SSC for photopic vision. Therefore, in the calculations, we used V2(λ) to denote the 2-degree and V10(λ) to denote the 10-degree systems, respectively. V’(λ) is the SSC for scotopic vision. Both V(λ) and V’(λ) are shown in Figure1. To compare the S/P ratio differences at each wavelength, we calculated the S/P ratios of monochromatic wavelengths for a full width at half maximum (FWHM) distance of 1 nm, from 380 to 780 nm, as shown in Figure2. The S/P ratio of an equal energy white (EEW) light calculated using 2-degree SSC is 2.261 and 10-degree is 2.071, and the difference is approximately 9%. It is clearly noted that the main differences appear in the short-wavelength range (<500 nm). Appl. Sci. 2019, 9, x FOR PEER REVIEW 2 of 9

CIE recommended mesopic vision models for road lighting [2] and published guidelines on how to use the model [12]. However, CIE did not recommend which SSC should be adopted for calculation, implying that either SSC type can be used in applications. In most of the published research studies and practical applications, the 2-degree SSC system has been used. Berman [13] reported S/P ratios of 16 typical light sources in the range from 0.23 to 2.47. Lita et al. [14] reported a light source with an S/P ratio of 2.56 at a correlated colour temperature (CCT) of 5368 K. Nizamoglu et al. [15] reported a white-light emitting diode (WLED) with an S/P ratio of 3.04 at the extremely high CCT of 45000 K. Guo et al. [16] proposed light-emitting diodes (LEDs) for CCTs from 2700 K to 45000 K with S/P ratios in the range from 1.83 to 3.84. Zan et al. [17] reported WLEDs that reached S/P ratios from 1.54 to 3.84, which corresponded to CCTs in the range from 2700 K to 7580 K. Li et al. [18] reported two-hump quantum-dot-based LEDs that reached an S/P ratio value of 2.09 at 4038 K, and a four-hump LED that reached an S/P ratio value of 3.08 at 5015 K. Zhang et al. [19] optimised hybrid white LEDs for high S/P ratios at varying CCTs that reached high colour rendition, and compared their differences. Despite the fact that in some studies it has not been mentioned which SSC was used, we rechecked the spectra in the literature and have confirmed that most studies have used 2-degree SSCs. To date, no study has compared the differences of the calculations of S/P ratios using these two SSCs. In practical applications, there are urgent needs for large-scale assessments of urban lighting [20] and nightscape lighting. However, there are some unsolved technical issues to assure photopic adaptation luminance [21,22], which may cause inaccuracy in assessments. In this study, we examined the differences of the calculated S/P ratios with two SSCs based on tests with more than 300 light sources [23] with CCTs from 2000 K to 8000 K, and with blackbody radiant (BR) light sources with CCTs from 10000 K to 45000 K. We calculated the mesopic luminance levels at photopic adaptation luminance levels from 0.005 to 5 cd/m2 based on the use of the two tested SSCs and compared the differences. Conversion coefficients between the S/P ratio and mesopic luminance calculated using 2- and 10-degree SSCs were proposed. The work may help elucidate the S/P ratios and mesopic luminance differences calculated with these two different SSCs in research studies, and provide convenient conversions in practical applications.

2. S/P Ratio of Monochromatic Spectra

The S/P ratio was calculated by dividing the LER in scotopic vision (LERs) by that in photopic vision (LERp). LERp was defined by Equation (1), LERs was defined by Equation (2), and the S/P ratio, which can also be expressed as a quotient in the form of Rsp [12], was defined with the use of Equation

(3). Herein, P(λ) represents the spectral power distribution (SPD) of the light source, Km is the

maximum photopic luminous efficacy of 683 lm/W at 555 nm, and Km is the maximum scotopic luminous efficacy of 1700 lm/W at 507 nm. Accordingly, V(λ) is the SSC for photopic vision. Therefore,

Appl.in the Sci. calculations2019, 9, 4471 , we used V2(λ) to denote the 2-degree and V10(λ) to denote the 10-degree systems3 of 10 , respectively. V’(λ) is the SSC for scotopic vision. Both V(λ) and V’(λ) are shown in Figure 1.

1.0 V'(l)

0.8 V2(l)

V10(l) 0.6

) l

(

P Appl. Sci. 2019, 9, x FOR PEER REVIEW0.4 3 of 9

Figure 1. Spectral sensitivity0.2 curves (SSCs) for scotopic vision and photopic vision using 2- and 10-degree systems. 0.0 To compare the S/P ratio380 differences480 at each580 wavelength,680 we780 calculated the S/P ratios of monochromatic wavelengths for a full widthWavelength at half maximum (nm) (FWHM) distance of 1 nm, from 380 to 780 nm, as shown in Figure 2. The S/P ratio of an equal energy white (EEW) light calculated using 2-degreeFigure SSC 1. Spectralis 2.261 sensitivityand 10-degree curves is (SSCs)2.071, and for scotopicthe difference vision is and approximately photopic vision 9%. using It is clearly 2- and noted that10-degree the main systems.differences appear in the short-wavelength range (<500 nm). (nm) (nm) 380 440 531.1 555 380 555 (nm) 100 660 780 100 440 533.5 555 660 780 100 R = 84.429 Rsp = 35.506 sp 2 degree 10 degree Rsp = 13.155 R = 72.586 10 sp 10 10

Rsp = 2.261 R = 2.071 Rsp = 1 R = 1 sp Rsp = 1 Considerable difference 1 sp 1 1

sp sp sp

R Equal to R R R sp Equal to Rsp 0.1 of EEW light 0.1 of EEW light 0.1 R = 0.026 Rsp = 0.026 sp

0.01 R =0.013 0.01 0.01 sp Rsp = 0.013 Minor difference 0.001 0.001 0.001 380 480 580 680 780 380 480 580 680 780 380 480 580 680 780 Wavelength (nm) (a) Wavelength (nm) (b) Wavelength (nm) (c) Figure 2. Scotopic/photopic ratios (R ) as a function of wavelength using (a) 2-degree and (b) 10-degree Figure 2. Scotopic/photopic ratios sp(Rsp) as a function of wavelength using (a) 2-degree and (b) SSCs. Some special Rsp values have been marked. The Rsp at 555 nm is equal to unity. The Rsp 10-degree SSCs. Some special Rsp values have been marked. The Rsp at 555 nm is equal to values at 531.1/533.5 nm are 2.261/2.085, and are equal to those for equal energy white (EEW) light unity. The Rsp values at 531.1/533.5 nm are 2.261/2.085, and are equal to those for equal inenergy the 2-degree white/ 10-degree(EEW) light cases. in the (c) Comparison2-degree/10- ofdegreeS/P ratios cases using. (c) C 2-omparison and 10-degree of S SSCs./P ratios There using is a2 considerable- and 10-degree diﬀerence SSCs in. the There short-wavelength is a considerable part, difference especially for in values the short<500-wavelength nm, and a minor part, diﬀerence in the long wavelength part. especially for values <500 nm, and a minor difference in the long wavelength part.

R 780 K 780 P( )V( )d Km 380 P()()l Vλ l d lλ λ m 380 LERp = (1) LERp  780R 780 (1) Pd()llP(λ)dλ 380 380 R 780 780 'K P( )V ( )d Km0 380 P(l ) Vλ '( l ) d lλ λ = m 380 0 LERs (2) LERs  780R 780 (2) Pd()llP(λ)dλ 380 380 R 780 780 'K P( )V ( )d Km0 380 P(l ) Vλ '( l ) d lλ λ S/P ratio = m 380 0 . . (3) SP/ ratio  780R 780 (3) KKm P()()lP V(λ l)V d( lλ)dλ m 380 380 3. Relation of S/P Ratios of Light Sources Using 2-degree and 10-degree SSCs 3. Relation of S/P Ratios of Light Sources Using 2-degree and 10-degree SSCs Monochromatic spectra calculated for the S/P ratios in Section2 are considered extreme conditions. In thisMonochromatic section, we investigated spectra calculated a set that contained for the S more/P ratios than in 300 S lightection sources. 2 are consideredMost of these extreme were commercialconditions. light In this sources. section, These we lightinvestigated source data a set were that presented contained in more the Illuminating than 300 light Engineering sources. SocietyMost of ofthese North were America commercial (IES-30-15. light 2015), sources. including These incandescent light source datalamps, were ﬂuorescent presented lamps, in the high-intensity Illuminating dischargeEngineering lamps, Society and of LEDs. North The America CCTs of (IES these-30- light15. 2015), sources including ranged incandescent from 2000 to 8000lamps, K. fluorescent The SPDs werelamps, normalised high-intensity to a maximum discharge intensity lamps, and of unity. LEDs. The The mean CCTs SPD of these is shown light sources in Figure ranged3. We from calculated 2000 to 8000 K. The SPDs were normalised to a maximum intensity of unity. The mean SPD is shown in Figure 3. We calculated S/P ratios of these light sources using both 2- and 10-degree SSCs. Figure 4 shows that the S/P ratio was calculated in the plot of the 10-degree against the 2-degree SSC system. The relation is fitted well by a linear function with a high coefficient of determination, R2 = 0.999. The slope of the linear function is approximately equal to the ratio of the EEW light, which is equal to 2.071/2.261  0.916. We can use 0.916 as the conversion coefficient for the transformation of the results calculated with the use of the 2- and the 10-degree SSCs. Appl. Sci. 2019, 9, 4471 4 of 10

S/P ratios of these light sources using both 2- and 10-degree SSCs. Figure4 shows that the S/P ratio was calculated in the plot of the 10-degree against the 2-degree SSC system. The relation is fitted well by a linear function with a high coefficient of determination, R2 = 0.999. The slope of the linear function is Appl. Sci. 2019, 9, x FOR PEER REVIEW 4 of 9 approximately equal to the ratio of the EEW light, which is equal to 2.071/2.261 0.916. We can use 0.916 as the conversion coefficient for the transformation of the results calculated with the use of the 2- and Appl. Sci. 2019, 9, x FOR PEER REVIEW 4 of 9 Mean SPD of 318 light sources with error bars the 10-degree SSCs. 1.0

0.8 Mean SPD of 318 light sources with error bars 1.0 0.6 0.8 0.4 0.6

Intensity (a.u.) 0.2 0.4 0.0

Intensity (a.u.) 0.2 -0.2 0.0380 430 480 530 580 630 680 730 780 Wavelength (nm) -0.2 380 430 480 530 580 630 680 730 780 Figure 3. Mean spectral power distribution (SPD)Wavelength of 318 (nm) light sources with error bars, as provided by the Illuminating Engineering Society of North America (IES-30-15). The corresponding colorimetric Figure 3. Mean spectral power distribution (SPD) of 318 light sources with error bars, as provided by parametersFigure 3. Mean were spectral as follows: power correlated distribution colour (SPD) temperatures of 318 light (CCT sources) 3628 with K, x error = 0.397, bars y, =as 0.384, provided CIE Raby the Illuminating Engineering Society of North America (IES-30-15). The corresponding colorimetric =the 87, Illuminating IES Rf = 86, andEngineering Rg = 100. Society of North America (IES-30-15). The corresponding colorimetric parameters were as follows: correlated colour temperatures (CCT) 3628 K, x = 0.397, y = 0.384, CIE Ra parameters were as follows: correlated colour temperatures (CCT) 3628 K, x = 0.397, y = 0.384, CIE Ra = 87, IES Rf = 86, and Rg = 100. = 87, IES Rf = 86, and Rg = 100.

3.0 S/P ratio Linear regression 2.5 Slope = 2.071/2.261 3.0 S/P ratio 2.071 Linear regression 2.0 2.5 Slope = 2.071/2.261 EEW light

1.5 2.071

in 10-degree 2.0

EEW light

sp Slope 0.94

R Slope 0.916 1.0 1.5 2.261

in 10-degree

sp Slope 0.94 y = 0.94x R Slope 0.916 0.5 2 1.0 R = 0.999 2.261 0.5 1.0 1.5 2.0 2.5 y = 0.943.0x 0.5 2 Rsp in 2-degree R = 0.999 0.5 1.0 1.5 2.0 2.5 3.0 . Figure 4. Rsp calculated in a plot of the 10-degree against the 2-degree SSC system. The red-dashed line Rsp in 2-degree isFigure a linear 4. regressionRsp calculated fitting in with a plot a slope of the of 10 0.94.-degree The blue against solid the line 2 is-degree a straight SSC line. system. with a The slope red of - 0.916,dashed which line is equalis a linear to the ratioregression of the R spfittingvalues with for EEW a slope light in of the 0.94. 10-degree The blue (2.071) solid and 2-degreeline is a (2.261) systems. straightFigure 4. lineRsp calculated with a slope in ofa plot 0.916 of, thewhich 10- degreeis equal against to the ratio the 2 of-degree the Rsp SSCvalues system. for EEW The light red- indashed the 10 -linedegree is a (2.071) linear andregression 2-degree fitting (2.261) with system a slopes. of 0.94. The blue solid line is a Due to the fact that there are no differences in the responses for scotopic vision, the S/P ratio straight line with a slope of 0.916, which is equal to the ratio of the Rsp values for EEW light difference is caused by the difference of LER . Although considerable differences may exist for S/P Duein the to 10 the-degree fact that (2.071) there and are 2 -nodegree difference (2.261)p s insystem the responsess. for scotopic vision, the S/P ratio ratios,difference the di isff erencescaused by for the mesopic difference luminance of LER valuesp. Although may not considerable be as considerable. difference Wes may should exist further for S/P examineratios,Due the the todifference mesopic the fact s luminancethat for theremesopic are values luminanceno difference transformed valuess in frommaythe responses photopicnot be as adaptationconsiderablefor scotopic luminance vision,. We should the levels S /furtherP ratio in mesopic vision states. examinedifference the is caused mesopic by lumin the differenceance value ofs transformedLERp. Although from considerable photopic adaptation difference luminances may exist levels for S in/P mesopicratios, the vision difference states. s for mesopic luminance values may not be as considerable. We should further 4. Mesopic Luminance Values at Different Photopic Adaptation Luminance Levels examine the mesopic luminance values transformed from photopic adaptation luminance levels in 4.mesopic MesopicFigure vision5 Lshowsuminance state thes. plot Values of the at 10-degreeDifferent versusPhotopic the A 2-degreedaptation luminance Luminance values Levels at the photopic adaptation luminance levels of 0.01, 0.03, 0.1, 0.3, 1, and 3 cd/m2. These data are well fitted (high Figure 5 shows the plot of the 10-degree versus the 2-degree luminance values at the photopic 4. Mesopic Luminance Values at Different Photopic Adaptation Luminance Levels adaptation luminance levels of 0.01, 0.03, 0.1, 0.3, 1, and 3 cd/m2. These data are well fitted (high R2 values)Fig withure 5 linear shows functions the plot. ofThis the implies 10-degree that versusphotopic the adaptation 2-degree luminance luminance values is the main at the factor photopic that influencadaptationes theluminance results. levels The slope of 0.01, increases 0.03, 0.1, from 0.3, 1.011, and to 3 2.23 cd/m with2. The photopicse data are adaptation well fitted luminance (high R2 valuesvalues) in with the linearrange functionsfrom 0.01. Thisto 3 impliescd/m2. For that photopic photopic adaptation adaptation luminance luminance values is the main <0.1 cd/mfactor2, thatthe differencesinfluences theare very results. small. The When slope the increases luminance from values 1.01 to>0.3 2.23 cd/m with2, the photopic plotted adaptationpoints are scattered. luminance values in the range from 0.01 to 3 cd/m2. For photopic adaptation luminance values <0.1 cd/m2, the differences are very small. When the luminance values >0.3 cd/m2, the plotted points are scattered. Appl. Sci. 2019, 9, 4471 5 of 10

R2 values) with linear functions. This implies that photopic adaptation luminance is the main factor thatAppl. inﬂuences Sci. 2019, 9, the x FOR results. PEER REVIEW The slope increases from 1.01 to 2.23 with photopic adaptation luminance5 of 9 values in the range from 0.01 to 3 cd/m2. For photopic adaptation luminance values <0.1 cd/m2, the Appl. Sci. 2019, 9, x FOR PEER REVIEW 2 5 of 9 diﬀerences are very small. When the luminance values0.06>0.3 cd/m , the plotted points are scattered. 0.020 0.01 cd/m2 0.03 cd/m2 0.05 0.016 0.06 2 2 0.0200.012 0.01 cd/m 0.04 0.03 cd/m 0.05

0.016) y = 1.01x 0.03 y = 1.05x 2 0.008 R2 = 0.999 0.04 R2 = 0.999 0.0120.004 0.02 0.004 0.008 0.012 0.016 0.020 0.02 0.03 0.04 0.05 0.06 ) y = 1.01x 0.03 y = 1.05x 2 0.008 0.42 0.16 0.1 cd/mR2 = 20.999 0.3 cd/mR2 = 20.999 0.004 0.020.38 0.140.004 0.008 0.012 0.016 0.020 0.02 0.03 0.04 0.05 0.06 0.42 2 2 0.160.12 0.1 cd/m 0.34 0.3 cd/m 0.38 0.140.10 y = 1.06x 0.30 y = 1.16x-0.04 R2 = 0.999 0.34 R2 = 0.997 0.120.08 0.26 0.08 0.10 0.12 0.14 0.16 0.26 0.30 0.34 0.38 0.42 0.101.3 y = 1.06x 0.303.5 y = 1.16x-0.04 2 2 2 2 1 cd/mR = 0.999 3.4 R 3= cd/m0.997 0.08Mesopic luminance in 10-degree (cd/m 1.2 0.26 0.08 0.10 0.12 0.14 0.16 3.30.26 0.30 0.34 0.38 0.42 1.3 3.5 1.1 2 3.2 2 1 cd/m 3.4 3 cd/m Mesopic luminance in 10-degree (cd/m 1.2 3.1 1.0 y = 1.32x-0.29 3.3 y = 2.23x-3.61 3.0 1.1 R2 = 0.989 3.2 R2 = 0.947 0.9 2.9 0.9 1.0 1.1 1.2 1.33.1 2.9 3.0 3.1 3.2 3.3 3.4 3.5 1.0 y = 1.32x-0.29 y = 2.23x-3.61 3.0 R2 =Mesopic 0.989 luminance in 2-degree (cd/mR2 = 20.947) 0.9 2.9 0.9 1.0 1.1 1.2 1.3 2.9 3.0 3.1 3.2 3.3 3.4 3.5 2 Figure 5. Mesopic luminance plots Mesopicof the luminance 10- against in 2-degree the (cd/m 2-degree) system s at the reference photopic adaptation luminance levels of 0.01, 0.03, 0.1, 0.3, 1, and 3 cd/m2. The reference Figure 5. Mesopic luminance plots of the 10- against the 2-degree systems at the reference photopic Figurephotopic 5. Mesopic adaptation luminance luminance plots of the level 10 - isagainst included the 2 -degree in the system 2-degrees at the SSC reference system. adaptation luminance levels of 0.01, 0.03, 0.1, 0.3, 1, and 3 cd/m2. The reference photopic adaptation photopicAccordingly, adaptation corresponding luminance photopic levels of adaptation 0.01, 0.03, 0.1,luminance 0.3, 1, and values 3 cd/m in the2. The 2-degree reference SSC luminance level is included in the 2-degree SSC system. Accordingly, corresponding photopic adaptation photopicsystem should adaptation be multiplied luminance with level the is conversion included coefficient in the 2- degreeRsp (2- degree)/ SSC system.Rsp (10 - luminance values in the 2-degree SSC system should be multiplied with the conversion coefficient Rsp Accordingly,degree). corresponding photopic adaptation luminance values in the 2-degree SSC (2-degree)/Rsp (10-degree). system should be multiplied with the conversion coefficient Rsp (2-degree)/Rsp (10- Figuredegree).Figure6 shows 6 shows the mesopicthe mesopic luminance luminance ratio estimated ratio estimated by dividing by dividing the mesopic the mesopic luminance luminance values ofvalues the 10-degree of the 10 SSC-degree by the SSC 2-degree by the SSC.2-degree The mesopic SSC. The luminance mesopic luminance ratio increases ratio as increases the photopic as the adaptationphotopicFigure luminanceadaptation 6 shows theincreasesluminance mesopic from increases luminance 1.000 in from scotopic ratio 1.000 estimated vision in scoto to 1.057picby dividingvision in photopic to the1.057 mesopic vision. in photopic Specifically, luminance vision. thevaluesSpecifically difference of the, ofthe 10 mesopic -differencedegree luminance SSC of bymesopic the ranged 2- degreeluminance from SSC. 0% ranged to The 5.7% mesopic from based 0% on luminance to the 5.7% calculations based ratio on increases withthe calculat the useas theions of thephotopicwith two the SSCs. adaptationuse of the two luminance SSCs. increases from 1.000 in scotopic vision to 1.057 in photopic vision. Specifically, the difference of mesopic luminance ranged from 0% to 5.7% based on the calculations with the use of the two SSCs. 1.08

1.07 1.057 1.08 1.06 1.053 1.071.05 1.046 1.057 1.039 1.061.04 1.053 1.051.03 1.046 1.039 1.028 1.04 1.02 1.021 1.031.01 1.011 1.028 1.021.00 1.021 1.000 1.01 1.011

Mesopic 10-degree and 2-dgree luminance ratio of 0.99 0.003 0.01 0.03 0.1 0.3 1 3 5 1.00 1.000Photopic adaptation luminance (cd/m2)

Mesopic 10-degree and 2-dgree luminance ratio of 0.99 0.003 0.01 0.03 0.1 0.3 1 3 5 Figure 6. Mesopic luminance ratio using 10- and 2-degree light source SSCs at photopic adaptation Figure 6. Mesopic luminance ratioPhotopic using adaptation 10- luminanceand 2- (cd/mdegree2) light source SSCs at photopic luminance values from 0.005 to 5 cd/m2. adaptation luminance values from 0.005 to 5 cd/m2. Figure 6. Mesopic luminance ratio using 10- and 2-degree light source SSCs at photopic 5. Examinationsadaptation luminance of BR Light values Sources from at0.005 Extremely to 5 cd/m H2igh. CCTs At higher CCTs, the differences of the S/P ratio calculated using two 2-degree and 10-degree SSC 5. Examinations of BR Light Sources at Extremely High CCTs systems tend to be greater. We invested BR light sources in the temperature range from 10000 to 45000At K,higher and resultsCCTs, theare differencesshown in Fig ofure the 7. S/P The ratio differences calculated increase using twowith 2 CCT,-degree and and they 10 are-degree over SSC10%. systems tend to be greater. We invested BR light sources in the temperature range from 10000 to 45000 K, and results are shown in Figure 7. The differences increase with CCT, and they are over 10%. Appl. Sci. 2019, 9, 4471 6 of 10

Appl.5. Examinations Sci. 2019, 9, x FOR of PEER BR LightREVIEW Sources at Extremely High CCTs 6 of 9 At higher CCTs, the differences of the S/P ratio calculated using two 2-degree and 10-degree Figure 8 shows the mesopic luminance ratio as a function of CCT and the photopic adaptation Appl.SSC Sci. systems 2019, 9, x tend FOR PEER to be REVIEW greater. We invested BR light sources in the temperature range from 100006 of 9 luminance. The mesopic luminance ratio increased as the CCT and photopic adaptation luminance to 45000 K, and results are shown in Figure7. The di fferences increase with CCT, and they are over increased. The maximum difference was approximately equal to 16%. Fig10%.ure Figure8 shows8 shows the mesopic the mesopic luminance luminance ratio ratio as a as function a function of of CCT CCT and and the the photopic photopic adaptation adaptation luminaluminance.nce. The The m mesopicesopic luminance luminance ratio ratio increase increasedd as as the the CCT CCT and and photopic photopic adaptation adaptation luminance luminance increasedincreased.. The The maximum maximum difference diff3.8erence was was approximate approximatelyly equal equal to to 16%. 16%. 0.90 2-degree 3.6 10-degree 3.8 0.900.89 3.4 2-degree 3.6 10-degree 3.2 0.890.88

sp R 3.4 3.0

0.87 in 10-degree and 2-degree 3.2 0.88 sp

R 2.8 3.0

0.870.86 in 10-degree and 2-degree

Ratio of 2.6 sp

R 2.8 10000 15000 20000 25000 30000 35000 40000 45000 CCT (K) 0.86

2.6 Ratio of

10000 15000 20000 25000 30000 35000 40000 45000 CCT (K) Figure 7. Rsp variations for CCTs in the range from 10000 to 45000K and ratio of Rsps in 10- and 2-degree SSC systems as a function of CCT. Figure 7. Rsp variations for CCTs in the range from 10000 to 45000K and ratio of Rsps in 10- and 2-degree

FigureSSC systems 7. Rsp variations as a function for ofCCTs CCT. in the range from 10000 to 45000K and ratio of Rsps in 10- and 2-degree SSC systems as a function of CCT. Mesopic luminance ratio 5 3

)

2 1.16 Mesopic luminance1.14 ratio 51 3 1.12

)

2 1.16 0.3 1.10 1.14 1 1.08 0.1 1.12 1.06 0.3 1.10 1.04 0.03 1.08 0.1 1.02 1.06

Photopic adaptation (cd/m luminance Photopic 0.01 1.00 1.04 0.0050.03 10000 15000 20000 25000 30000 35000 40000 45000 1.02

Photopic adaptation (cd/m luminance Photopic 0.01 CCT (K) 1.00 0.005 Figure 8. Mesopic luminance10000 ratio15000 using20000 10-25000 and30000 2-degree35000 40000 SSCs45000 as a function of CCT and photopic Figureadaptation 8. Mesopic luminance. luminance ratio usingCCT 10- (K) and 2-degree SSCs as a function of CCT and photopic adaptation luminance. 6. Examination of a Real Lighting Scene Figure 8. Mesopic luminance ratio using 10- and 2-degree SSCs as a function of CCT and 6. ExaminationWe measured of a luminanceReal Lighting distribution Scene of a road in Shenzhen using a typical phosphor converted photopic adaptation luminance. LEDWe type. measured The road luminance is 140 m distribution in length and of 16a road m in in width, Shenzhen as shown using in a Figuretypical9 phosphor. We used converted a spectral 6.LEDimage Examination type. radiance The roadof colorimeter a Riseal 140 L ightingm EVERFINE in length Scene and SIRC-2000 16 m in to width measure, as shown the luminance in Figure distributions 9. We used and a spectral spectral imageinformation radiance of light colorimeter sources. EVERFINE The colorimeter SIRC- was2000 set to at measure 60 m far the away luminanc and 1.5e m distributions in height in and each We measured luminance distribution of a road in Shenzhen using a typicalS phosphorP converted spectralmeasurement. information The of luminaire light sources. was measuredThe colorimeter at a CCT was ofset 3650at 60 K,m far and away the and/ ratio 1.5 m in in theheight 2- and in LED10-degree type. The SSCs road were is 1.33140 m and in 1.28,length respectively. and 16 m in The width average, as shown luminance in Figure was approximately9. We used a spectral 1 cd/m 2. imageeach measurement. radiance colorimeter The luminaire EVERFINE was measured SIRC-2000 at a to CCT measure of 3650 the K, and luminanc the S/eP distributionsratio in the 2 - andand The referred paradigm was a typical road lighting application at moderate CCT and luminance2 spectral10-degree information SSCs were of 1.33 light and sources. 1.28, respectively The colorimeter. The averagewas set atluminance 60 m far awaywas approximately and 1.5 m in height 1 cd/m in. Thlevels.e referred We transformed paradigm was the a photopic typical road adaptation lighting luminanceapplication to at mesopic moderate luminance CCT and inluminance a point-by-point levels. eachmanner. measurement. Figure 10 showsThe luminaire the luminance was measured and the corrected at a CCT mesopicof 3650 K, luminance and the S distributions./P ratio in the There 2- and are We transformed the photopic adaptation luminance to mesopic luminance in a point-by-point2 10810-degree luminance SSCs were points 1.33 in and the road,1.28, r whichespectively form. The the luminanceaverage luminance distribution was contours.approximately The di1 ﬀcd/merence. Thmanne referreder. Figure paradigm 10 shows was the a typical luminance road lightingand the applicationcorrected mesopic at moderate luminance CCT and distributions. luminance levelTheres. Weare 810 transform luminanceed the points photopic in the road adaptation, which form luminance the luminance to mesopic distribution luminance contours. in a point The difference-by-point mannbetweener. Figurethe corrected 10 shows mesopic the luminance luminance and using the bothcorrected the 10 mesopic- and 2- degreeluminance SSC sdistributions. of these 810 pointsThere areis approximately 810 luminance 3% points on average in the road, as ,shown which inform Figure the luminance11 . distribution contours. The difference between the corrected mesopic luminance using both the 10- and 2-degree SSCs of these 810 points is approximately 3% on average, as shown in Figure 11 . Appl. Sci. 2019, 9, 4471 7 of 10

Appl. Sci. 2019, 9, x FOR PEER REVIEW 7 of 9 Appl. Sci. 2019, 9, x FOR PEER REVIEW 7 of 9 between the corrected mesopic luminance using both the 10- and 2-degree SSCs of these 810 points is approximatelyAppl. Sci. 2019, 9, x 3%FOR on PEER average, REVIEW as shown in Figure 11. 7 of 9

Figure 9. Measurements conducted in a typical road in Shenzhen with a length of 140 m and Figurea width 9. Measurementsof 16 m. conducted in a typical road in Shenzhen with a length of 140 m and a width of 16 m. Figure 9. Measurements conducted conducted in ain typical a typical road road in Shenzhen in Shenzhen with a lengthwith a of length 140 m of and 140 a width m and of 16a width m. of 16 m. Luminance (cd/m2) Luminance (cd/m2) Luminance (cd/m2) Luminance (cd/m2) 140 Luminance (cd/m2.072) 140 Luminance (cd/m2.072) 140 Luminance (cd/m2.072) 140 Luminance (cd/m2.072) 140 2.071.74 140 2.071.74 140 2.071.74 140 2.071.74 120 1.741.40 120 1.741.40 120 1.741.40 120 1.741.40 120 Luminance (cd/m1.401.072) 120 Luminance (cd/m1.401.072) 120 Luminance (cd/m1.401.072) 120 Luminance (cd/m1.401.072) 140100 1.072.070.74 140100 1.072.070.74 140100 1.072.070.74 140100 1.072.070.74 0.40 0.40 0.40 0.40 100 0.741.74 100 0.741.74 100 0.741.74 100 0.741.74 1.400.07 1.400.07 1.400.07 1.400.07 12080 0.40 12080 0.40 12080 0.40 12080 0.40 0.071.07 0.071.07 0.071.07 0.071.07 80 80 80 80 100 0.74 100 0.74 100 0.74 100 0.74 Length (m) 60 Length (m) 60 Length (m) 60 Length (m) 60 0.40 0.40 0.40 0.40

Length (m) 60 Length (m) 60 Length (m) 60 Length (m) 60 0.07 0.07 0.07 0.07 8040 8040 8040 8040 40 40 40 40

Length (m) 6020 Length (m) 6020 Length (m) 6020 Length (m) 6020 20 20 20 20 400 400 400 400 0 4 8 12 16 0 4 8 12 16 0 4 8 12 16 0 4 8 12 16 0 0 0 0 0 4 Width8 (m)12 16 (a) 0 4 Width8 (m)12 16 (b) 0 4 Width8 (m)12 16 (c) 0 4 Width8 (m)12 16 (d) 20 20 20 20 Width (m) (a) Width (m) (b) Width (m) (c) Width (m) (d) 0 0 0 0 Figure 010.4 Luminance8 12 16 distributions0 4 and8 12 corrected16 mesopic0 4 luminance8 12 16 distributions0 4 8. (a12) Luminance16 Figure 10. Luminance distributions and corrected mesopic luminance distributions. (a) Luminance Figuredistributions 10. LuminanceWidth measured (m) distributions(a) using the Width2 and-degree (m) corrected SSC(b) system mesopic, (Widthb luminance) c (m)onversion(c) distributions to luminanceWidth. ( a(m) ) distributionsLuminance(d) distributions measured using the 2-degree SSC system, (b) conversion to luminance distributions using distributionsusing the 10 - measureddegree SSC using system the, ( c2)- degreecorrected SSC mesopic system luminance, (b) conversion using the to luminance2-degree system distributions, and (d ) theFigure 10-degree 10. Luminance SSC system, distributions (c) corrected and mesopic corrected luminance mesopic using luminance the 2-degree distributions system, and. (a) ( dLuminance) corrected usingcorrected the 10 mesopic-degree luminance SSC system using, (c) c theorrected 10-degree mesopic system. luminance using the 2-degree system, and (d) cmesopicdistributionsorrected luminance mesopic measured luminance using using the 10-degreeusing the 2 the-degree 10 system.-degree SSC systemsystem., (b) conversion to luminance distributions using the 10-degree SSC system, (c) corrected mesopic luminance using the 2-degree system, and (d) corrected mesopic luminance using1.04 the 10-degree system. 1.04

1.03 1.04 1.03

1.02 1.03 1.02

1.01 1.02 1.01

1.00 1.01 0 90 180 270 360 450 540 630 720 810 1.00Mesopic luminance ratio of 10-degree and 2-degree 0 90 180 270 360 450 540 630 720 810 Mesopic luminance ratio of 10-degree and 2-degree Luminance points Luminance points 1.00 Figure 11. Mesopic luminance ratios0 of90 10-degree180 270 and360 2-degree450 540 630 SSCs720 of810 810 luminance points. The mean

Figure 11. Mesopic luminanceMesopic luminance ratio of 10-degree and 2-degree ratios of 10-degree and 2-degree SSCs of 810 luminance points. The Figure 11. Mesopic luminance ratios of 10-degree and 2-degree SSCs of 810 luminance points. The ratiomean is ratio 1.03, is which 1.03, yieldswhich ayields difference a difference of ~3%. ofLuminance ~3%. points mean ratio is 1.03, which yields a difference of ~3%. 7. Discussion and Conclusion 7. DiscussionFigure 11. Mesopicand Conclusion luminance ratios of 10-degree and 2-degree SSCs of 810 luminance points. The 7. DiscussionInmean many ratio applications,and is 1.03 Conclusion, which yields the 2-degree a difference SSC of was ~3% used. for the calculation of the LER and S/P ratio. In many applications, the 2-degree SSC was used for the calculation of the LER and S/P ratio. However,In many in someapplications, applications, the 2-degree a large SSC visual was field used was for required, the calculation and a 10-degreeof the LER SSC and wasS/P ratio. more 7.However, Discussion in someand C applications,onclusion a large visual field was required, and a 10-degree SSC was more However,reasonable. in Wesome analy applications,sed the difference a large visualbetween field these was two required, SSCs based and aon 10 an-degree investigation SSC was of more more In many applications, the 2-degree SSC was used for the calculation of the LER and S/P ratio. reasonable.than 300 light We sources analys edand the extremely difference high between CCT BR these sources. two SSCs based on an investigation of more thanHowever, 300 light in somesources applications, and extremely a large high visual CCT BR field sources. was required, and a 10-degree SSC was more reasonable. We analysed the difference between these two SSCs based on an investigation of more than 300 light sources and extremely high CCT BR sources. Appl. Sci. 2019, 9, 4471 8 of 10 reasonable. We analysed the difference between these two SSCs based on an investigation of more than 300 light sources and extremely high CCT BR sources. Comparison of the S/P ratios calculated using monochromatic light sources yielded considerable differences in the short-wavelength region (<500 nm). Light sources operating at high CCT values will likely cause considerable differences. In the very short-wavelength range (<430 nm), the S/P ratio increased abruptly. However, human eyes are not sensitive to changes in this region, and the precision of SSCs cannot guarantee the precision of the extremely high S/P ratio. In the mesopic vision range, the differences in the mesopic luminance values of the 300 light sources calculated with the two SSCs increased as the photopic adaptation luminance values increased. In the scotopic vision cases, the same SSC V’(λ) was applied and there were no differences. In photopic vision cases, the average difference reached a value of ~6%. For some practical road lighting applications with luminance values less than 1 cd/m2 and with moderate CCTs at approximately 4000 K, the difference of the corrected mesopic luminance values based on the two different SSCs studied herein was approximately 3%. For high CCT values (>6000 K), light sources, and high luminance levels (>1 cd/m2), the difference may reach 5%. Accordingly, as the differences of the two SSCs increase, the CCTs of the light sources also increased. The differences were relatively small and with values <5% for light sources operating with CCTs in the range from 2000 to 8000 K. For CCTs > 10000 K, the error at some mesopic luminance levels reached values >10%. In some research work, CCT light sources with extremely high S/P ratios have been reported. The S/P ratios for these light sources calculated using the two SSCs are so different that they require further analyses. For some special applications, like dashboard illumination in driving cabins and landscape lightings, blue LEDs might be used. Accordingly, considerable differences may be evoked. However, a few light sources applied in functional lighting operate at extremely high CCTs. Generally, the average difference of the S/P ratio is approximately 8% for light sources operating at CCTs in the range from 2000 to 8000 K. The mesopic luminance values, which are related to both the photopic adaptation luminance and CCT, do not yield considerable S/P ratio differences. For functional lighting, the photopic adaptation luminance is the main reason for the evoked differences. In summary, several valuable inferences were drawn based on the findings of this study. Firstly, in most conditions, to transform the S/P ratio calculated using the 2- and 10-degree SSC outcomes, a conversion coefficient equal to 0.916 should be used. Secondly, in most practical applications, the difference of the calculated mesopic luminance using 2- and 10-degree SSCs was approximately in the range of 3%–5%, and the conversion coefficients ranged from 1.03 to 1.05. Thirdly, for applications with CCTs > 10,000 K, and for luminance values >1 cd/m2, the differences may be >10%, and mesopic luminance should be calculated using suitable SSCs. Most urban lighting scenes, such as road lighting and tunnel lighting, are in the mesopic vision range, and in the nightscape lighting. Most of these lighting scenes require large visual field, while others may require small visual fields to observe. Applying the suitable SSC to reach more precise assessments could help improve the lighting design, yield a smaller energy consumption, and reduce the risk of potential light pollutions. There are some limitations associated with this study. Firstly, it is difficult to determine a very precise photopic adaptation luminance for complex luminance distributions because adaptation luminance may be related to luminance distributions, eye movements, and surrounding luminance effects [21]. Secondly, the adaptation luminance determined from 2-degree and 10-degree visual fields should be different, because the visual angles and adaptation fields [12] are different. The CIE TN 007:2017 recommends the adaptation field to be taken as the design area, and the adaptation luminance as the average luminance of the adaptation field. This is not applicable to designing suitable adaptation fields for each observation point when assessing a large-scale urban lighting scene. There is no feasible way at present to apply a suitable adaptation luminance for the entire road. Therefore, in this work, we only consider the SSC as the factor influencing photopic adaptation luminance and adopt a local photopic adaptation luminance for each observation point. Appl. Sci. 2019, 9, 4471 9 of 10

Author Contributions: Conceptualisation, Jianhua. Ding. and Qi. Yao.; methodology, Qi. Yao.; validation, Lei. Jiang.; formal analysis, Lei. Jiang.; investigation, Jianhua Ding.; writing—original draft preparation, Jianhua Ding. and Qi Yao.; writing—review and editing, Qi. Yao.; funding acquisition, Qi.Yao. Funding: This research was funded by the National Natural Science Foundation of China (NSFC) (grant number 61605125) and Beijing Advanced Innovation Center for Future Urban Design, Beijing University of Civil Engineering and Architecture (Project No. UDC2018010411). Conﬂicts of Interest: The authors declare no conﬂict of interest.

References

1. CIE 115:2010: Lighting of Roads for Motor and Pedestrian Traffic; CIE: Vienna, Austria, 2010. 2. CIE 191:2010: Recommended System for Mesopic Photometry based on Visual Performance; CIE: Vienna, Austria, 2010. 3. CIE. Commission Internationale de l’Eclairage Proceedings; Cambridge University Press: Cambridge, UK, 1926. 4. CIE Proceedings (1964) Vienna Session (Committee Report E-1.4.1); Bureau Central de la CIE: Paris, France, 1963; Volume B, pp. 209–220. 5. Stockman, A.; Sharpe, L.T. Human cone spectral sensitivities: a progress report. Vis. Res. 1998, 38, 3193–3206. [CrossRef] 6. Shaw, M.Q. Evaluating the 1931 CIE color matching functions. In Proceedings of the AIC, The University of Hertfordshire, UK, 25–30 May 1997. 7. Sharpe, L.T.; Stockman, A.; Jagla, W.; Jagle, H. A luminous efficiency function, v*(l), for daylight adaptation. J. Vis. 2005, 5, 948–968. [CrossRef][PubMed] 8. Boyce, S.F.P.; Richards, M. Road lighting and energy saving. Lighting Res. and Technol. 2009, 41, 245–260. [CrossRef] 9. Akashi, Y.; Rea, M.S. Peripheral detection while driving under a mesopic light level. J. Illuminating Eng. Soc. 2002, 31, 85–94. [CrossRef] 10. Akashi, Y.; Rea, M.S.; Bullough, J.D. Driver decision making in response to peripheral moving targets under mesopic light levels. Lighting Res. Technol. 2007, 39, 53–67. [CrossRef] 11. Stockman, A.; Sharpe, L.T. Into the twilight zone: the complexities of mesopic vision and luminous efficiency. Ophthalmic Phys. Optics 2006, 26, 225–239. [CrossRef][PubMed] 12. CIE TN 007:2017: Interim Recommendation for Practical Application of the CIE System for Mesopic Photometry in Outdoor Lighting; CIE: Vienna, Austria, 2017. 13. Berman, S.W. Energy efficiency consequences of scotopic sensitivity. J. IES 1992, 1, 3–14. [CrossRef] 14. Lita, A.; Washington, A.L.; van de Burgt, L.; Strousse, G.F.; Stiegman, A.E. Stable efficient solid-state white-light-emitting phosphor with a high scotopic/photopic ratio fabricated from fused CdSe–silica nanocomposites. Adv. Mater. 2010, 22, 3987. [CrossRef][PubMed] 15. Nizamoglu, S.; Erdem, T.; Demir, H. High scotopic/photopic ratio white-light-emitting diodes integrated with semiconductor nanophosphors of colloidal quantum dots. Opt. Lett. 2010, 36, 1893. [CrossRef][PubMed] 16. Guo, Z.Q.; Shih, T.M.; Lu, Y.J.; Gao, Y.-L.; Zhu, L.-H.; Chen, G.-L.; Zhang, J.-H.; Lin, S.-Q.; Chen, Z. Studies of scotopic/photopic ratios for color-tunable white light-emitting diodes. IEEE Photon J. 2013, 5, 8200409. 17. Zan, L.L.; Lin, D.Y.; Zhong, P.; He, G. Optimal spectra of white led integrated with quantum dots for mesopic vision. Opt. Express 2016, 24(7), 7643. [CrossRef][PubMed] 18. Li, F.; You, L.; Nie, C.; Zhang, X.J.; Li, H.; Gu, X.; Huang, Y.; Li, Q. Quantum dot white light emitting diodes with high scotopic/photopic ratios. Opt. Express 2017, 25(18), 21901. [CrossRef][PubMed] 19. Zhang, C.W.; Xiao, L.C.; Zhong, P.; He, G. Photometric optimization and comparison of hybrid white LEDs for mesopic road lighting. Appl. Opt 2018, 57, 4665–4671. [CrossRef][PubMed] 20. Yao, Q.; Wang, H.B.; Uttley, J.; Zhuang, X. Illuminance reconstruction of road lighting in urban areas for efficient and healthy lighting performance evaluation. Appl. Sci. 2018, 8, 1646. [CrossRef] 21. Uchida, T.; Ayama, M.; Akashi, Y.; Hara, N.; Kitano, T.; Kodaira, Y.; Sakai, K. Adaptation luminance simulation for CIE mesopic photometry system implementation. Lighting Res. Technol. 2016, 48, 14–25. [CrossRef] Appl. Sci. 2019, 9, 4471 10 of 10

22. Gao, C.; Xu, Y.; Wang, Z.; Melosa, M.; Pointer, M.; Luo, M.R.; Xiao, K.; Li, C. Improved computation of the adaptation coeﬃcient in the CIE system of mesopic photometry. Opt. Express 2017, 25, 18365. [CrossRef] [PubMed] 23. Illuminating engineering society. IES method for evaluating light source colour rendition. IES 2015, 15–30.

© 2019 by the authors. Licensee MDPI, Basel, Switzerland. This article is an open access article distributed under the terms and conditions of the Creative Commons Attribution (CC BY) license (http://creativecommons.org/licenses/by/4.0/).