Figure 1

The chromaticity coordinates for which spectra where successfully optimised and strategies of assessing the non-visual quality of the metameric spectra. (A) We used a uniform grid of 561 chromaticity coordinates along the Planckian locus (2700 K to 7443 K ± Duv 0 to 0.048 with Duv-steps of 0.003) as optimisation targets for spectral optimisation. Depending on the LED luminaire’s channel configuration, the number of successfully optimised spectra per chromaticity target differs. The 6-channel LED luminaire, for example, was unable to yield spectra for chromaticity targets with higher Duv values at CCTs between 2700 and 3000 K. The blue scatter points in the CIExy-2° colour space highlight the chromaticity coordinates for which metameric spectra were successfully optimised. (B) For each of the blue points in the CIExy-2° colour space (left panel), numerous metameric spectra were available. When using a single chromaticity coordinate at 5552 K (Duv of 0.021, red point in the left panel) as an optimisation target for the 11-channel luminaire, 386 spectra were available (see panel 1). We defined spectra whose chromaticity coordinates were inside the tolerance range of \(\Delta u',\Delta v'\le 0.001\) (to the optimisation target) as metameric spectra (see panel 1). From the remaining spectra the upper and lower limit of \({E}{}_{mel}^{D65}\) can be classified, representing the melanopic tuning range \(\Delta {E}{}_{mel }^{D65}\) for the single chromaticity coordinate (see panel 2). Each chromaticity coordinate target is characterised with the respective \({E}{}_{mel, Min}^{D65}\), \({E}{}_{mel, Max}^{D65}\), \(\Delta {E}{}_{mel}^{D65}\), \({\gamma }{}_{mel,Min}^{D65}\), \({\gamma }{}_{mel,Max}^{D65}\), \(\Delta {\gamma }{}_{mel}^{D65}\) and its melanopic Michelson Contrast \(C_{M}\). (C) When considering the chromaticity coordinates for one CCT (left panel), the upper (blue points) and lower limit (black points) of the melanopic DER (\({\gamma }{}_{mel,Min}^{D65}\), \({\gamma }{}_{mel,Max}^{D65}\)) can be plotted and analysed as a function of the Duv steps, making possible to state which Duv distance to Planck is ideal for spectral optimisation. The largest or lowest melanopic EDI (and respective melanopic DER) values of a CCT (across all Duv values) at a photopic illuminance of 250 lx will be denoted as \({\widehat{E}}{}_{mel, Min}^{D65}\) (brown point), \(\widehat{E}{}_{mel, Max}^{D65}\) (red point). As the melanopic EDI values correlate with the melanopic DER, these key points can also be expressed with \(\widehat{\gamma}{}_{mel,Min}^{D65}\) and \(\widehat{\gamma }{}_{mel,Max}^{D65}\). The largest melanopic tuning range of a CCT will be denoted as \(\Delta \widetilde{E}{}_{mel}^{D65}\) or for a generalised description as \({\Delta \widetilde{\gamma}}{}_{mel}^{D65}\). (D) To compare the non-visual metrics across the different CCT steps, we contracted the values for each Duv value to one axis with blue points as the upper limit of the melanopic DER tuning range and black points for the lower limit. The red point states the \(\widehat{\gamma}{}_{mel,Max}^{D65}\) and accordingly the \(\widehat{E}{}_{mel, Max}^{D65}\) value (see panel C)).