Astronomy tools · Dark skies
SQM to Bortle and NELM.
Enter an SQM reading (mag/arcsec²), a naked-eye limiting magnitude, or a Bortle class. Get the other two, the formula behind them, and a warning when the numbers sit on a class edge.
Dark sites get described three ways: the number on a Sky Quality Meter, the faintest star you can see, and a Bortle class. The usual pointer for converting between the first two is the Unihedron conversion page, which holds two one-line equations and no explanation. This page runs those same equations, then places the answer on the Bortle table and says plainly where that table and the equations disagree.
Short answer: A reading of 21.00 mag/arcsec² gives a naked-eye limiting magnitude of about 6.1 by the Unihedron equation. The Wikipedia Bortle table puts 21.00 in class 4 by its SQM column and 6.1 in class 4.5 by its NELM column, so the two columns already disagree by half a class.
What the three numbers measure
A Sky Quality Meter reads how bright the sky is per patch of sky, in magnitudes per square arcsecond. Bigger means darker, and the scale is logarithmic, so a one-point gain is about 2.5 times less light. Naked-eye limiting magnitude (NELM) is something a person does: the faintest star you can see. The US National Park Service puts it simply, “NELM is the magnitude of the faintest star we can see in the sky with naked eyes,” and gives anchors: 6.6 is near pristine under average conditions, 7.0 needs good seeing and dark adaptation, and a number lower than 6.3 usually means degraded sky. The Bortle class is a 1 to 9 label devised by amateur astronomer John Bortle and published in Sky & Telescope in February 2001.
The two equations
The Unihedron page, written by K. Fisher (revised August 2006) and crediting Schaefer’s 1990 paper on telescopic limiting magnitude, gives NELM = 7.93 − 5 log10(10(4.316 − SQM/5) + 1) and the inverse, SQM = 21.58 − 5 log10(10(1.586 − NELM/5) − 1). The page is titled as a conversion between a V-band NELM system and a B-band MPSAS system, which is a reminder that these are fitted relationships, not physical identities. The tool also shows a second estimate from Crumey’s fit on Wikipedia’s limiting magnitude page, with a field factor of 2, so you can see how far two published models differ for the same sky. At 21.00 they differ by about a quarter of a magnitude.
Where the Bortle table fits
The classes use the approximate SQM and NELM columns on Wikipedia’s Bortle scale page: class 1 is 21.76 to 22.0 and NELM 7.6 to 8.0, class 4 is 20.8 to 21.3 and 6.3 to 6.5, class 4.5 is 20.3 to 20.8 and 6.1 to 6.3, class 8 is below 18.00, and class 9 has no SQM value at all on that page. The page says the classes “correspond closely” with SQM and NELM, not exactly, and notes a 2014 study (Crumey) suggesting Bortle may have overstated what typical observers see, with a NELM of 6.0 to 6.5 more often cited than the 7.1 and up listed for class 2. When your SQM and NELM land in different classes, trust neither more than the other.
A warning about very dark readings
The National Park Service notes that the SQM “does not reliably measure the sky brightness when the sky is darker than 21.5 mag/arcsec²,” and treats 22.0 as the darkest part of pristine skies. If your reading is above 21.5 the result below is flagged, because the meter itself is the weak link.
Worked example
Take an SQM reading of 21.00. The exponent is 4.316 − 21.00/5 = 0.116, and 100.116 is 1.306. Add 1 to get 2.306; log10 of that is 0.363; times 5 is 1.815; and 7.93 − 1.815 = 6.12 for the NELM. Feeding the unrounded 6.1155 back through the inverse returns 21.00. On the Bortle table, 21.00 sits between 20.8 and 21.3, so class 4 by SQM, while 6.12 falls in the 6.1 to 6.3 band, so class 4.5 by NELM. Crumey’s fit for 21.00 gives 5.85. These are worked from the published equations, not measurements of a real site.
Frequently asked questions
How do I convert SQM to Bortle class?
Find your SQM reading in the SQM column of a Bortle table: 21.3 to 21.6 is class 3, 20.8 to 21.3 is class 4, 19.25 to 20.3 is class 5, and so on. The boundaries are approximate, so a reading near an edge could sit in either class.
What is the formula for SQM to NELM?
NELM = 7.93 − 5 log10(10^(4.316 − SQM/5) + 1), from the Unihedron conversion page. The inverse is SQM = 21.58 − 5 log10(10^(1.586 − NELM/5) − 1).
Why do my SQM and NELM give different Bortle classes?
The Bortle table’s two columns come from different sources and only correspond closely. Near a class boundary a small difference in either number moves the class, so treat the class as a rough label.
Can the SQM read a perfect Bortle 1 sky?
Not reliably. The National Park Service says the instrument does not reliably measure skies darker than 21.5 mag/arcsec², so a reading near 22 is best read as very dark, not as a precise figure.
Is the NELM result what I will see?
It is an estimate for an average observer. Eyesight, dark adaptation, altitude of the star and sky transparency all move the real limit, and the Wikipedia Bortle page notes a study suggesting the table overstates what typical observers see.
Sources and how it is calculated
Retrieved
The two equations are the published Unihedron ones. The Bortle class comes from the lower bound of each band in the Wikipedia table. The second NELM estimate is Crumey’s fit with a field factor of 2. Pages were read on 26 October 2026.
- Unihedron, NELM (V) to MPSAS (B) conversion calculator: both equations, K. Fisher, rev. 8/2006
- Wikipedia, Bortle scale: class table with approximate SQM and NELM columns; Bortle, Sky & Telescope, February 2001
- Wikipedia, Limiting magnitude: Crumey’s fit between sky brightness and limiting magnitude
- US National Park Service, Metrics Guide to Night Skies Report: NELM anchors and the SQM limit near 21.5 mag/arcsec²
Runs entirely in your browser. Nothing you type is sent or stored. Not affiliated with any forum, software maker or publisher named on this page.
An estimate, not a measurement
Real limiting magnitude depends on the observer and the night. Use this to compare sites and readings, not to certify a sky.