Watch the video

Human visual performance in low-light environments has been shown to improve under whiter light sources (those with S/P Ratio > 1). Until now, lighting professionals have had no way to quantify this effect. IES document TM-12-12 provides us with a system to revise our exterior lighting calculations to show this improved visibility.


  • The eye has two types of photoreceptors, Rods and Cones.
    • The rods are more numerous than cones and scattered across the retina, primarily outside of the Macula (Fovea is the center of the Macula). Rods are more sensitive to light, but do not provide color information, and are active during very low-light situations.
    • The cones are concentrated in the fovea. Cones are active under medium and high light levels, are necessary for on-axis viewing, and are responsible for color perception.
  • Vision during the daytime or indoors under average light levels corresponds to the Photopic luminous efficiency function, with spectral sensitivity peaking at 555 nm wavelength (yellow-green). This is “cone vision,” where the cones are most active.
  • Vision in darkness (moonless night, no exterior lighting) corresponds to the Scotopic luminous efficiency function, with peak spectral sensitivity at 507 nm (blue-green). This is pure rod vision.
  • Exterior lighting occurs in the area between Photopic and Scotopic vision, called the “Mesopic” range. Here both rods and cones are active.
  • The IES defines Mesopic vision to occur at adaptation luminance levels less than 5 cd/m2 and greater than .03 cd/m2 (IES)
  • Generally, lighting design photometric calculations are based on the lumen and derived quantities (candela, footcandle, lux), which are based on the photopic luminous efficiency function. This means our lighting calculations are optimized for photopic conditions, or interior lighting.
  • Using mesopic multipliers as detailed in TM-12-12, we can correct our calculations for low-light applications like outdoor lighting.
  • The spectral performance of a light source can be expressed in terms of its Scotopic luminous flux divided by its Photopic luminous flux, or S/P Ratio. Whiter light sources typically have S/P ratios >1.

Implementation in AGi32:

  1. Assign the correct S/P Ratio to the light source in AGi32 (Define Luminaire dialog). S/P ratios should be obtained from the luminaire or source manufacturer. If not available, some typical examples are provided in TM-12-12, Annex C.
  2. An object must be placed as the pavement (or ground) surface with accurate reflectance value, as the metric is luminance based.
  3. Place the calculation points using AGi32’s Automatic Placement command with “Calculation Type” of “Photopic/Mesopic.” AGi32 will automatically create four grids of calculation points: Photopic Luminance, Mesopic Luminance, Photopic Illuminance, Mesopic Illuminance. Mesopic Illuminance is visible by default.
  4. Calculate using the Full Radiosity method. All grids are calculated.


The “Effective Luminance” (Mesopic Luminance) is computed from the Photopic Luminance weighted by the S/P Ratio of the source. The Mesopic Illuminance at each point can then be back-calculated as the luminance is diffuse (not direction sensitive). The results in terms of Mesopic Illuminance are displayed by default (any of the grids can be made visible using Project Manager) and can be readily compared to the Photopic Illuminance numbers.

As mesopic effects are magnified for lower light levels, it is typical to see the minimum value rise appreciably under sources with S/P ratio >1. This, in turn, affects the uniformity ratios of Avg/Min and Max/Min in a positive way. The inverse is true for sources with S/P ratio <1. Mesopic results are actually worse than Photopic and uniformity ratios degraded. For S/P ratio = 1, there is no mesopic effect.

The AGi32 implementation of TM-12-12 is valid for exterior environments such as parking areas and residential streets, where speeds are less than 25 mph (40 kph). This is due to the nature of the visual task. The effect of spectrum on vision at mesopic light levels is in off-axis and peripheral vision, both of which play a larger role when the observer is stationary or moving slowly. In addition, the TM-12-12 calculation methodology has not been adopted for roadway lighting, where there may be other factors involved in the computation of adaptation luminance such as observer location, roadway pavement properties and veiling luminance. Lighting Analysts is awaiting guidance from the IES for implementation of this next step.