Reading the numbers
What the numbers mean, and the few ways they are easy to misread.
mpsas: larger is darker
Sky brightness is measured in magnitudes per square arcsecond (mpsas). The scale runs backwards from what you might expect: a larger number is a darker sky. About 21.5 is a dark rural sky; 18 is a bright suburban one. Each step of 1.0 is about 2.5 times more or less light.
The meter is accurate to about ±0.10 mpsas, so smaller differences than that are not real.
Bortle classes
Readings are labelled with the Bortle dark-sky class their value falls in:
| Class | mpsas | Sky |
|---|---|---|
| 1 | 21.76 and darker | excellent dark-sky site |
| 2 | 21.60 – 21.76 | typical truly dark site |
| 3 | 21.30 – 21.60 | rural sky |
| 4 | 20.80 – 21.30 | brighter rural sky |
| 4.5 | 20.30 – 20.80 | semi-suburban / transition |
| 5 | 19.25 – 20.30 | suburban sky |
| 6 | 18.50 – 19.25 | bright suburban sky |
| 7 | 18.00 – 18.50 | suburban / urban transition |
| 8–9 | brighter than 18.00 | city and inner-city sky |
The colours in the first column are the ones SQM History uses. They follow the light-pollution maps many astronomers know, softened: greys for Classes 1 and 2, then blue, green, yellow-green, yellow, orange and red, and near-white for Classes 8–9. The value is always written as well, so nothing depends on colour alone.
A class here describes a reading, not your site. Bortle classes describe a site from what an observer sees across the whole sky. The meter sees a patch about 20° wide, and moonlight, cloud, twilight and snow all change what it reads.
Special readings
| 0.00 (saturated) | The meter is too brightly lit to measure: daylight, or a light shining on it. It is shown as SATURATED, never counted, and never drawn as zero. |
|---|---|
| Stale (S) | At a dark site the sensor can need tens of seconds per reading. Asked again too soon, it repeats its last value and marks it stale. Stale readings are kept and shown but not counted. If you see many, make Sample every longer. |
Three averages
| Mean | The ordinary average of the readings. The headline figure. |
|---|---|
| Median | The middle reading. Not pulled about by a passing car’s headlights. |
| Mean brightness (flux) | The average amount of light, turned back into mpsas. It is always a little lower (brighter) than the mean, and differs most on nights that changed a lot. |
"Darkest" is not "best"
The darkest reading is the darkest the sky was measured to be. It is not proof of a clear night: thick cloud at a dark site can read darker than a clear sky, because cloud blocks starlight too. The application never guesses cloud or seeing from brightness alone.
What is drawn, and what is counted
| The chart | Runs from sunset to sunrise, so you can see how much of the night was recorded. |
|---|---|
| The statistics | Count only astronomical night, when the sun is more than 18° below the horizon. This part is shaded on the chart. |
Far enough north in midsummer, the sun never gets 18° below the horizon. Then there is no astronomical night, and the application says so rather than counting something else.
In SQM History you can choose another Interval: nautical darkness, the whole session, or the older rule of counting only readings above 17.00. That last one is there to match figures published by earlier tools. It favours dark readings by design, so its averages come out darker.