How a star map is calculated
24 October 2026
For each map we turn the place, date and local time you choose into the position of every star in the sky, then project the visible half of the sky onto a disc. The stars come from the Hipparcos and XHIP catalogues, and accuracy is about a tenth of a degree, far finer than a poster can show.
1. The data: real stars
The positions and brightness of the stars were measured by the European Space Agency’s Hipparcos satellite (1997). We use the XHIP compilation (F. Anderson and C. Francis, 2012) and Olaf Frohn’s d3-celestial file, which draws from it the 5,044 stars down to magnitude 6, together with the outlines of the 88 constellations of the International Astronomical Union. The sources and licences are listed on the Sources and credits page.
2. The place: a latitude and a longitude
When you type a city, the suggestions come from GeoNames and the search for a precise place from OpenStreetMap. We keep its latitude and longitude. Latitude decides how high the stars stand in the sky, longitude which part of the sky is passing overhead.
3. The time: from time zone to universal time
The time you enter is the local time of the chosen place. We apply that place’s time zone and any daylight saving in force on that date, using the IANA time zone database, then convert it to universal time. The same 10:00 pm in Paris is not the same universal time in January and in August.
4. Sidereal time and precession
From universal time and longitude we compute local sidereal time, which tells which part of the celestial sphere is crossing the meridian. The catalogues give positions for the year 2000; but the Earth’s axis drifts slowly, by about one degree every 72 years (the precession of the equinoxes). We therefore correct each position for the chosen date, using the standard formulas of Jean Meeus, Astronomical Algorithms.
5. The projection onto a disc
Each star is converted to a height above the horizon and a direction (azimuth), then the visible half of the sky is projected onto a disc with a stereographic projection. North is at the top and east on the left, as when you look at the sky. Stars below the horizon are not drawn.
Accuracy and limits
The calculation works for any place and any date from 1900 to 2100. Accuracy is about a tenth of a degree. Stars fainter than magnitude 6, the planets and the Moon are not drawn: the map is an artistic rendering of real data.
What if you do not know the exact time?
Enter an approximate time. The sky turns by about 15 degrees per hour: a one-hour error rotates the map by that angle without changing which stars are visible. Failing that, midnight is a good choice, and the time can be left off the poster.
Frequently asked questions
Is the sky on a star map really accurate?
Yes: the position of every star is calculated for the chosen place, date and local time, taking the time zone, daylight saving and the precession of the equinoxes into account, with an accuracy of about a tenth of a degree.
Which dates does the map work for?
Every date from 1900 to 2100, from a wedding last year to a birth several decades ago.
Where do the stars come from?
From the Hipparcos (European Space Agency) and XHIP catalogues, through the d3-celestial file that gathers the 5,044 stars down to magnitude 6.
Why does the time change the map?
Because the Earth rotates: the sky turns by about 15 degrees per hour around the place of observation.