The solar neighbourhood, sixty thousand stars deep.
Every star in this view has real coordinates — Hipparcos and Gaia parallaxes converted to Cartesian parsecs, with blackbody colours derived from each star's B-V index. The Sun sits at the origin. The most famous stars — Sirius, Vega, Betelgeuse, Polaris, Alpha Centauri — are picked out as a second layer with hover and click; the rest is a faithful background of the solar neighbourhood out to about three thousand light-years.
Three.js · HYG database · 59k stars · Bloom post-processing
drag to rotate · scroll to zoom · hover a bright star for its name and class
data · HYG Database v40 (astronexus, MIT) · Hipparcos + Gaia DR3
Real positions
Each dot sits at its actual Cartesian position in parsecs, derived from RA, Dec and parallax. Parallax — the apparent shift of a nearby star against the background as Earth orbits the Sun — is how astronomers measure distance directly. Gaia measures parallaxes for over a billion stars to microarcsecond precision; HYG merges the Gaia data with the older Hipparcos catalogue for the bright stars Gaia oversaturates.
Real colours
Star colour comes from the surface temperature, which we read from the B-V colour index in the catalogue. Hot O-type stars at 30,000 K are blue-white, cool M dwarfs at 3,000 K are deep red. The mapping uses the Tanner-Helland fit to Mitchell Charity's blackbody table — the same algorithm photo editors use to set the colour temperature of an image.
Magnitude as size
Apparent magnitude is a logarithmic brightness scale where each step of one corresponds to a factor of 2.512 — five steps double-doubles to a hundredfold. Each star's point size in the renderer follows the same scale, so the visual hierarchy matches what your eye would see in a perfectly dark sky.
What's missing
The view stops at three thousand parsec. The Sun is one of about two hundred billion stars in the Milky Way, but only a few million are within that radius — and we plot the brightest sixty thousand. To see the disc of the Galaxy, the bar, the bulge, you would need a wider cut and a much larger dataset. Phase 2 of this exercise extends to that scale.