Lighting : Lighting April 2016 - Vol 36 Issue 2
40 LIGHTING MAGAZINE | April/May 2016 April/May 2016 | LIGHTING MAGAZINE 41 then sends signals to a network of luminaires, which generate the desired SPDs and direct the light to the appropriate locations. Though physical implementation of such a system would be complex, modern sensors, such as the Microsoft Kinect, are capable of real-time tracking of changing stimuli  and LEDs can be modulated faster than human vision can temporally resolve . Spectrally tuneable light sources have been developed, using LEDs [4-5]. The use of even spectrally narrower components, such as lasers , would allow increased in precision in SPD generation. Before technical implementation of such a system can be realistically considered, several fundamental questions must be addressed. In this project, we investigate the SPD properties that most effectively minimise light absorbed by objects, while rendering the colour appearance of illuminated objects naturally. METHODS In this investigation, SPDs, tailored to individual reflective objects, were mathematically optimised to minimise Figure 3. Relative power as a function of wavelength for test SPDs (black solid line) and the incandescent reference illuminant (dotted line). Reflectance as a function of wavelength (grey solid line) for a saturated yellow object (sample 5). light absorption and render the object colours naturally . The coloured objects studied were the 15 reflective samples used in the colour quality scale (CQS) . For the colorimetric calculations, the appearance of each test sample was determined, relative to a reference illuminant, in the CIE 1976 L*a*b* colour space . The reference illuminants were a standard incandescent lamp and an equal- energy radiator. For the single peak optimizations, each iterative process began with one wavelength within the visible spectrum (380 nm to 780 nm), designated as the “starting point.” In each subsequent computation, optical power was added, in 1.0 nm intervals, to each side of the starting point, to create spectral bandwidth. The bandwidth ranged from 1.0 nm to 401 nm, in odd numeric intervals. All combinations of starting points and bandwidth were iteratively simulated, and colour differences (ΔE*ab) were recorded for each test SPD, where ΔE*ab=1.0 is considered to be a barely recognisable difference . Simulations assumed that the observer had an adapted white point, corresponding to the chromaticity of the reference illuminant. Colour difference calculations were accompanied by calculations of the energy consumption of the test SPD relative to the reference illuminant. In these computations, the spectral power of the test SPD was adjusted, so that illuminated objects would have the same luminance as when lit by the reference illuminant. A scaling variable was calculated for each test SPD to ensure that the amount of light reflected by each object was the same for the test SPD and reference illuminant. When the reference illuminant was an equal energy radiator, the scaling variable corresponded to the peak value of the test SPD. When the reference was incandescent, this variable was used to scale the SPD, but the overall shape of the SPD was dictated by the spectrum of an incandescent lamp. For the two-peak SPD optimisations, the methods were identical, except each SPD had two separate starting points. The test SPDs were developed such that the two starting points could have different bandwidths, but the spectral power associated with each starting point could not overlap. An example of a two-peak test SPD, reference illuminant SPD, and reflectance factor of one of the coloured test samples are shown in Figure 3. RESULTS AND DISCUSSION For each test sample, all SPDs that would render the sample’s colour indistinguishably (ΔE*ab<1.0) from the reference illuminant were recorded. They were then sorted by the amount of energy they would save, relative to illumination with the reference illuminant. The SPD that yielded the greatest energy savings, and met the colour appearance requirement, was noted for each coloured sample. The averaged results for the single peak and two-peak optimisations are shown in Table 1 and Table 2, respectively. The results indicate that the energy consumed by lighting could be substantially decreased if lighting systems illuminated objects with SPDs that minimised absorption, while yielding the same colour appearance and luminance of objects as traditional white References  National Research Council, Assessment of Advanced Solid State Lighting. Washington: The National Academies Press, 2013.  R.A. Newcombe, S. Izadi, O. Hilliges, D. Molyneaux, D. Kim, A. Davison, P. Kohi, J. Shotton, S. Hodges, and A. Fitzgibbon, “KinectFusion: Real-time dense surface mapping and tracking,” 10th IEEE International Symposium on Mixed and Augmented Reality (ISMAR), pp. 127-136, October 2011.  E.F. Schubert, Light-Emitting Diodes, Cambridge: Cambridge University Press, 2006.  C.C . Miller, Y. Ohno, W. Davis, Y. Zong, and K. Dowling, “NIST spectrally tunable lighting facility for color rendering and lighting experiments,” Light & Eng., vol. 17(4), pp. 57-61, 2009.  K. Yuan, H. Yan, and S. Jin, “LED-based spectrally tunable light source with optimised fitting,” Chin. Opt. Lett., vol. 12(3), pp. 032301, 2014.  A. Neumann, J.J. Wierer, W. Davis, Y. Ohno, S.R.J. Brueck, and J.Y. Tsao, “Four-color laser white illuminant demonstrating high color-rendering quality,” Opt. Express, vol. 19(104), pp. A982-A990, 2011.  D. Durmus and W. Davis, “Optimising Light Source Spectrum For Object Reflectance,” Opt. Express, vol. 23(11), pp. A456-A464, 2015.  W. Davis and Y. Ohno, “Colour quality scale,” Opt. Eng., vol. 49(3), pp. 033602, 2010.  International Commission on Illumination, Colorimetry, Publication no. 15, Vienna: Central Bureau of the CIE, 2004.  R.W.G . Hunt and R.M . Pointer, Measuring Colour, 4th ed. West Sussex: John Wiley & Sons, Ltd, ch. 3, 2011. light sources. The results show an effect of the reference illuminant on energy savings. This is simply a consequence of the low luminous efficacy of radiation (LER) of incandescent lamps. These SPDs contain a large amount of power in the very long wavelengths, to which the human visual system is relatively insensitive. The proposed method of lighting would use SPDs that contain little or no power in the wavelengths that contribute the least to photometric measures of intensity. Future research will further investigate issues associated with the design and implementation of advanced lighting systems that minimise the amount of light absorbed by illuminated objects. In particular, experiments will be conducted that assess the perceived naturalness and attractiveness of real objects illuminated by absorption- minimizing SPDs. Research will also investigate the spatial resolution that would be required to illuminate individual objects within complex architectural spaces. Table 1. Summary of single peak test SPD characteristics when optimised for energy savings, as a function of the two reference illuminants, when target colour difference ΔE*ab <1.0. Values are the average of all the reflective samples’ highest energy saving condition. Reference Illuminant Starting Point (nm) Bandwidth (nm) Max. Energy Saving (%) Equal-energy 588 263 35 Incandescent 561 266 48 Table 2. Summary of two-peak test SPD characteristics when optimised for energy savings, as a function of the two reference illuminants, when target colour difference ΔE*ab <1.0. Values are the average of all the reflective samples’ highest energy saving condition. Reference Illuminant Starting Point 1 (nm) Bandwidth 1 (nm) Starting Point 2 (nm) Bandwidth 2 (nm) Max. Energy Saving (%) Equal-energy 468 42 575 52 52 Incandescent 477 60 588 69 62 Future research will further investigate issues associated with the design and implementation of advanced lighting systems that minimise the amount of light absorbed by illuminated objects.
Lighting February 2016 - Vol 36 Issue 1
Lighting June 2016 - Vol 36 Issue 3