Variability class

Eclipsing binaries

Eclipsing binaries don't actually change their light output. They orbit so that, from our line of sight, each star periodically hides the other — and the shape of those dips reads out the geometry of an entire stellar system.

2,637,124 starsEW · EA · EB

Phaseφ 0.00 · Δm 0.00

the star at this phase

W UMa type: continuous light, two near-equal minima. Algol type: flat between sharp eclipses.

What it is

A pair of stars bound in orbit, oriented edge-on enough that each passes in front of the other once per cycle. The deeper drop — the primary eclipse — happens when the hotter star is hidden; the shallower secondary half an orbit later. Because it is pure geometry, the light curve repeats once per orbit, so its period is the orbital period; for contact systems a periodogram often finds half of it (see below).

The physics

Eclipse depths encode the ratio of surface brightnesses; eclipse widths encode the stellar radii relative to the orbit; the spacing between primary and secondary reveals orbital eccentricity. A contact binary shares a common envelope, so its light varies continuously and its two minima are nearly equal. A detached pair sits quietly out of eclipse and drops sharply only during the brief eclipses.

VSX types

Subtypes

W UMa type

Continuous variation with near-equal minima; the components are nearly or fully in contact. Periods typically 0.2–0.8 d.

Algol type

Nearly constant light between eclipses, which are sharp and often unequal; detached or semi-detached (Algol itself is semi-detached). Periods from hours to years.

β Lyrae type

Continuous light variation from ellipsoidal components, with clearly unequal minima; usually semi-detached or near-contact.

Method

Finding the period

A contact binary shows two near-equal minima per orbit, so a floating-mean GLS fit latches onto the P/2 harmonic. The archive doubles it to the orbital period unless folding at the found period already shows two unequal minima. Algol-type systems (EA) are nearly flat between eclipses, where a sinusoid sees almost nothing, so varchive runs BLS on one light curve per eclipsing candidate (never on Gaia's) and prefers it for Algols. A BLS box or an Algol-like peak is doubled when its odd and even eclipses differ in depth; a detached pair with two equal eclipses stays at half its orbit. Every eclipser's epoch is then refined from our own fold to the primary minimum — never trusted from VSX — and used to predict the next eclipse.

Pitfalls

The top GLS peak for an EW is almost always P/2; never adopt it blindly.
A shallow secondary can vanish into the noise, making a genuine binary look like a single-dip transit.
A factor-of-two difference from the VSX period means one of the two is the half-period; unequal minima in the fold at the longer period show that it is the orbit.
A detached pair whose two eclipses are equal folds identically at half its orbit; photometry cannot tell the two apart, so the archive stays at the half period.
Faint, brief eclipses push the magnitude distribution to positive skew and high kurtosis — exactly the eclipse hint that selects a star for BLS.

Archive stars

Examples

ASASSN-V J080423.71-434822.9

EW · contact pair — GLS finds P/2; the archive doubles it to 0.62 d

#10000000474670 →

ASASSN-V J030643.65-753726.6

EA · a 5.3-day Algol-type eclipser, its period found by BLS

#10000002122083 →

ASASSN-V J074317.21-075440.9

EB · a β Lyrae–type pair, continuously varying

#10000002141040 →

Browse eclipsing binaries

2,637,124 stars in search →Populations