Research paper

  • Hello,

    You might find this paper interesting enough to add it to your references database:

    «Evaluating Potential Spectral Impacts of Various Artificial Lights on Melatonin Suppression, Photosynthesis, and Star Visibility.

    • Martin Aubé , Johanne Roby, Miroslav Kocifaj
      Published: July 5, 2013»


    In this paper, we introduced three new indices that can be used to characterize the SPD characteristics of any lighting device. These indices have been designed to allow a quick estimation of the potential impact of different lamp spectra on melatonin suppression, photosynthesis and star visibility. Indices have also been designed to separate the impact of the shape of the SPD from other factors, such as illuminance levels or the angular photometry of the lamp.

    We used the indices to compare different lighting technologies in term of their spectral impacts. In particular we found that LPS, phosphor-converted amber LEDs and LED 2700 K-filtered lamps have a lower potential impact on melatonin suppression in comparison to HPS. LPS, LED 5000 K-filtered and LED 2700 K-filtered lamps show a lower impact on photosynthesis compared to HPS. Only LPS, phosphor-converted amber LEDs and LED 2700 K-filtered lamps are better than or equivalent to HPS in restricting the impact on star visibility. We also showed that atmospheric scattering under clear or cloudy skies generally reduces the value of the indices in comparison to direct light, except for metal halide lamps for which the effect is opposite.

    During daytime, if we want to obtain the highest spectral impact, we should favor metal halide for increased melatonin suppression, while phosphor-converted amber LEDs and metal halide lighting is the best for stimulating photosynthesis.

    It is important to realize that the spectral indices introduced here should not be considered alone in making a complete evaluation of the impact of a given lamp installation. As an example, a lamp with low MSI (low relative blue content) can impact significantly on melatonin suppression when the lamp illuminance is so high that the absolute blue flux becomes important. Basically, a global evaluation can be obtained by multiplying the index by the illuminance.

    We hope that these indices will form some kind of standardization in characterization of lighting technologies, and that the lighting industry will provide such indices in the same way as they currently provide CCT or CRI, for example. Finally we consider that these indices will favor the design of new environmentally and health-friendly light devices, by providing a means of quickly evaluating the relative potential impact of a device at the design stage.»*

    Have a nice day,


  • We like Aube's work and have used the same original reference in our photosynthesis action spectrum.

    Aube's MSAS action spectrum seems to be derivative of the Rea model (2005), which combined Brainard's 2001 work with Thapan et al's (two different studies, by different groups, published around the same time).

    This combination of two studies (in the Rea model) creates a "notch" that nobody has documented in real life: notably it predicts that 4000K sources suppress less melatonin than 3000K sources! There is really no literature showing this in the same kind of light source, so we don't think it is a very good prediction. (Also, both studies were done with dilated pupils, which makes them hard to translate to polychromatic sources with freely-adapting pupils.)

    So, this is why we don't use it and for now just use the simple melanopic response, not because it is "better" (though it is for many things), but because it predicts fewer things that are hard to explain.


  • Thanks! This is really hardcore stuff for a mere mortal programmer ;-)

    Working towards my own lamp, one research paper at a time (