A star isn't one color. It emits a continuous spectrum, brightest at one wavelength and falling off on either side — close to what physics calls a blackbody curve. The peak of that curve depends on temperature: hot stars peak in the ultraviolet, cool stars peak in the infrared, and the visible band sits somewhere in between.
To turn that curve into a single number, photometers measure brightness through two standard filters defined by Harold Johnson and William Morgan in 1953:
- B (blue) — peaks near 445 nm, in the blue.
- V (visual) — peaks near 551 nm, in the yellow-green where the eye is most sensitive.
The color index is simply the magnitude difference: B − V.
Magnitudes count backwards — smaller numbers mean brighter — so a hot star, which pumps out more blue light than visual, has a smaller (brighter) B than V. Subtract: negative B−V. A cool red star is the reverse: dim in B, bright in V → positive B−V. The further from zero, the more extreme the temperature.
Why Vega is zero
The B−V scale was anchored to Vega (T ≈ 9,600 K) by historical accident — Johnson & Morgan picked a few bright A0V stars and defined their average B−V as 0.00. The whole rest of the catalog is calibrated against that one anchor. Modern photometric systems (like AB or Gaia G_BP − G_RP) use absolute physical zeros, but the Vega-based B−V is so entrenched that it survives in every observer's tool belt.