By  Dr B. A. Clark

Photometric trial in connection with a scientific paper by McFadden, Jones, Schoemaker et al. (2014) on obesity and exposure to light at night

B. A. Clark, 2 August 2015


To determine room illuminances corresponding to photometric ranges defined by 1. Ability to read newsprint; 2. Seeing hand or seeing across room; or 3. Not to see a hand to below absolute threshold.


Hagner illuminance meter reading to 0.1 lux

Unihedron Sky Quality Meter (SQM) reading to fainter than 23.3 stellar magnitudes/square arc second (fainter than 5.2 x 10-5 cd/m2)

A cobra head MV streetlight was recently replaced by a T5 fluorescent lamp in its luminaire in the Banyule street where the writer resides. The main bedroom of the writer’s house has windows facing the streetlight, which is about 10 m distant and 6⁰ below eyeheight inside the window. At night, the measured vertical illuminance at the bedroom window was 0.35 lux. (The Hagner meter used was showing values of 0.3 and 0.4 alternately, so the mean was recorded.) Practically all of this light is from the relatively new T5 streetlight.

At the top of the stairwell and landing outside the bedroom door, the window there is close to and in the same plane as the bedroom windows. The landing window also has 0.35 lux from the streetlight. With the drapes open, this illumination caused the ceiling (shaded from light directly coming from the streetlight) to have an SQM reading of 19.12 mag/sq arc sec, equivalent to 2.43 x 10-3 cd/m2. If its surface were a white lambertian diffuser, the illuminance required to produce this luminance would have been 2.43 x 10-3 x 2π, ie 15.3 mlux. Given that the walls and ceiling are light grey with a reflectance of ca 40%, the actual illuminances required would have been about 2.5 times greater. When the drapes were closed, the ceiling luminance dropped to 9.4 x 10-5 cd/m2, equivalent to 5.9 x 10-4 lux falling on a white diffuser. The total transmittance of the closed drapes is the ratio of the two luminances, approx. 3.9%.

Photometrically, the bedroom is roughly comparable with the stair landing space. In the bedroom, the curtain edges are better sealed against light leakage and the ceiling there with blinds closed is darker than 23.3 mag/sq arc sec (5.2 x 10-5 cd/m2), so the transmittance of the closed drapes is < 2.1%.

Visibilities of newsprint, hands and room were noted during these measurements. Additional observations were found necessary to define the thresholds with sufficient accuracy. These were done in an under-stairs cupboard in which fine control of low levels of illumination could be achieved by having the door slightly open.


Reading newsprint text was not possible at 0. 3 lux and 0.5 lux (directly measured values, not calculated). Slow reading was possible at 0.7 lux, and 1.0 lux made the task easier. A hand at arm’s length could be seen in positive contrast against darker parts of the room when its illuminance was as small as 0.05 lux. Less light was required to be able to see across the room, down to about 0.01 lux. A hand in shade could be seen in negative contrast against a matt grey wall lit by 1 mlux.


The ranges defined by the McFadden study for the illuminance within the bedroom thus appear to be:

  1. Brighter than 0.5 lux.
  2. From 0.5 lux down to 0.01 lux.
  3. Dimmer than 0.01 lux.

In general, bedroom walls will have a mean illuminance less than that of outdoor illumination shining on the outside of the bedroom window. But in the worst case, somebody sleeping in a position where an external streetlight is shining directly on them through an uncurtained window will be subject to as much as about 90% of the external normal illuminance provided by the streetlight.


It is highly likely that the difficulty for ordinary individuals to make the required visibility judgements would have resulted in some miscategorisation of bedrooms in the McFadden study. As with pooling the results over a range of illumination within categories, this would have the effect of weakening the association between illumination and obesity. In other words, the actual association is likely to be stronger than that found.

In the present case the writer’s bedroom with the drapes open would be within Category 2 of the McFadden study, near its upper limit. The installed street light is arguably a health hazard, despite the fact that it complies with current Australian Standards.

Many houses in Banyule have windows of habitable rooms closer than 10 m from streetlights and the windows are commonly lower than the streetlights and consequently in brighter parts of the luminaire’s output. In these cases, it is likely that such rooms with the drapes or blinds opened would be categorised as within Category 1. Note that it is surprisingly common for people to have inefficient blinds or drapes or they fail to deploy them fully, particularly if they also use translucent curtains. Furthermore, with the streetlights at or above the window level as is nearly always the case, any direct light that does enter the room could shine on the faces of occupants while they are in bed.

All of the above factors add weight to the need to limit incident stray light at night in any vertical plane at windows of habitable rooms. A practicable limit would be 0.1 lux. Lower values might well be justifiable.


Additional information just to hand:


The Lancet Oncology

Volume 17, No. 1, p23–24, January 2016

Melatonin: resetting the clock of cancer progression?

David J Pinato

Justin Stebbing


© 2016 Elsevier Ltd. All rights reserved.

To view the full text, please login as a subscribed user or purchase a subscription. Click here to view the full text on ScienceDirect.


Right from its first, unsteady steps 4 million years ago, the evolution of mankind has been marked by the rhythmic, incessant ticking of a silent, internal clock. A repetitive synchronised pendulum swing forcing any living organism to obey the rules of an ancestral, supreme biorhythm. A measure to preserve the especially precious state called homeostasis that each individual tissue within a multicellular organism is forced to maintain within its own biological equilibrium. Life or death, growth or decline, acceleration or hindrance are among the messages carried across every second from each individual cell to another by signalling molecules like hormones, a precious molecular link between the surrounding environment, ruled by changes in daylight, seasons and eras, and the human body, troubled by the challenges of health and illness.


Visit ScienceDirect to see if you have access via your institution.