| Constellation | Brightest star | Altitude | Direction | Transits in |
|---|
| Star | Mag | RA | Dec | Altitude | Direction |
|---|
The Earth turns once on its axis in 23 hours 56 minutes 4 seconds, not 24 hours. A solar day is longer because during one rotation the Earth has also moved a degree or so along its orbit, and it must turn that bit further to bring the Sun back to the same place. The stars are not chasing the Sun, so they keep the shorter day.
The consequence is the most useful thing there is to know about the night sky: a star rises about four minutes earlier every night — two hours a month, and a full twenty-four hours over a year. That is why the constellations are seasonal, why Orion belongs to winter evenings in the northern hemisphere and is a summer morning object, and why the same view comes back to the same clock time one year later.
Local sidereal time is simply which right ascension is overhead. It is the one number that says what the sky looks like: a star transits — crosses due south at its highest — at the moment local sidereal time equals its right ascension. The readout above shows it, and you can watch it gain on the clock.
A map of the ground has north up and east to the right, because you are looking down at it. A whole-sky map has north up and east to the left, because you are looking up. Hold the disc above your head with north pointing north and it matches what is there; hold it in front of you like a road map and everything is mirrored. Every planisphere ever printed shares the convention, and it is the commonest reason a beginner decides a star chart must be wrong.
The horizon view does not have this problem, which is most of the reason it exists. Face south and it draws what is in front of you the way you are standing: west on the right, east on the left, ground at the bottom. Nothing to hold over your head and nothing to mentally flip — it is a photograph of that part of the sky rather than a map of all of it.
A star never sets if its declination is greater than 90° − your latitude, and never rises if it is further south than your latitude − 90°. From London at 51.5°N, everything north of +38.5° is up all night, every night — the Plough, Cassiopeia, Cepheus — and everything south of −38.5° never appears at all, which is why the Southern Cross is a holiday constellation for most of Europe. The band between those two limits is what gives the sky its seasons, and the latitude slider is the quickest way to see it: the further from the equator you go, the more of the sky is permanently on and permanently off.
At 51.5°N on 15 January at 22:00, local sidereal time is about 5h 40m. So the stars near right ascension 5h 40m are due south and at their highest — which is precisely Orion's belt, at RA 5h 36m.
Alnilam, the middle star of the belt, has declination −1.2°, so it transits at an altitude of 90 − 51.5 − 1.2 = 37.3° — about a third of the way up the sky. Polaris meanwhile sits at 51.5° due north all night, every night, because the altitude of the pole is your latitude. That is the oldest navigation trick there is, and it still works.
Come back at the same clock time a month later and the whole sky has slid two hours west: the belt is past the meridian and heading down, and whatever was two hours behind it is due south instead.
Three steps, all of them plain arithmetic. The date and time become a Julian date; the Julian date and your longitude become local sidereal time; and sidereal time with a star's right ascension and declination become an altitude and a bearing. Everything drawn on this page is those last two numbers, projected two different ways.
The whole-sky disc is azimuthal equidistant. Distance from the centre is proportional to the angle down from straight up, so the rim is the horizon, halfway out is 45° up, and a ruler laid across the chart measures degrees directly. It stretches shapes near the rim — constellations low down look pulled sideways — which is the price of getting every altitude right on one picture.
The horizon view is gnomonic, which is what a camera does: straight lines in the sky stay straight, and shapes near the centre are undistorted. That is why it looks like what you would actually see. The cost is that it cannot show much at once — push the field of view past about 100° and the corners stretch badly, which is the same reason a very wide camera lens bulges at the edges. Two projections, because no single one can be honest about both the whole sky and the shape of one part of it.
The catalogue is 104 of the brightest stars with 14 constellation figures drawn between them. It was checked before it was used, by measuring the angular separations it implies — Orion's belt end to end, the three sides of the Summer Triangle, the Pointers to Polaris, both axes of the Southern Cross — because a catalogue cannot be derived from anything, and a transposed digit in a coordinate looks exactly like a correct one.
No, and it is not trying to be. Positions are J2000 catalogue coordinates with no correction for precession, nutation, aberration or refraction, so a star can be out by up to about half a degree — the width of the Moon. That is invisible when you are finding a constellation and useless when you are aligning a mount.
Not here. Planets move against the stars, so they need an ephemeris rather than a fixed catalogue, and a chart that put them in the wrong place would be worse than one that leaves them out. The Sun & Moon tool carries the arithmetic for those two.
Because of where you are standing. The chart draws everything down to the magnitude you choose; a city centre typically shows magnitude 3 and brighter, a suburb 4, genuinely dark countryside 6 or more. Drag the faintest-star slider down to match what you can really see and the chart becomes a much better map of your own sky.
Because this is a catalogue of the brightest stars, not a complete one. A constellation covers an area of sky containing hundreds of stars; what is charted here are the ones you can actually pick out, which for some figures is seven and for others is one. The count is stated on every constellation so it is never implying otherwise.
A spherical Earth, an observer at sea level with a flat, unobstructed horizon, and J2000 positions for 104 bright stars in 14 figures — not all 88 constellations, and not every star in the ones it does draw. No precession, nutation, aberration, parallax, proper motion or atmospheric refraction: all of them matter at the arc-minute level and none of them matter for finding Orion. Times are your device's local time, and the date control assumes that clock is right.
Releases in which this page changed, newest last. Derived from the archived copy of every release, not from notes written afterwards — so it reflects what actually shipped. Site-wide passes are left out; they are in the full changelog.