Apparent brightness and intrinsic luminosity

Begin with what the observation actually measures

A stellar image begins as a detector measurement: events or accumulated charge distributed across pixels and exposures. Extracting a source signal requires treatment of background, detector response, exposure and calibration. A large pixel value is not itself a direct measurement of the star’s total power.

Received flux is energy arriving per unit time and collecting area, often in a stated wavelength interval. Its SI unit can be watts per square metre. Spectral flux densities describe how that received signal is distributed with wavelength or frequency and require the corresponding extra unit.

Luminosity is power emitted by a source, measured in watts. A band luminosity includes only a specified portion of the spectrum; bolometric luminosity includes electromagnetic radiation over the complete spectrum. The wavelength limits belong to the quantity, not merely to a footnote.

Intrinsic describes a property assigned to the source rather than the amount received at our telescope. It does not mean immutable: a star’s total luminosity can vary with time. Its angular emission pattern can also be anisotropic, as considered below; that direction dependence is distinct from its total power. The observation epoch and adopted treatment of variability still matter.

Distance connects the quantities through assumptions

The preceding distance lessons established the transparent isotropic relation F_bol = L_bol/(4πd²). The surface of a sphere centred on the source grows with d², so the same emitted power is distributed over a larger area. This relation is a radiation model with stated conditions.

Consequently, a high-luminosity star can have a smaller received flux than a lower-luminosity star if the distances or intervening paths differ. Apparent brightness alone does not rank their intrinsic powers. A distance estimate and an appropriate treatment of attenuation are needed for that inference.

For a matched energy-defined passband, an analogous relation can connect band flux and band luminosity when the geometry and propagation assumptions apply. A detector’s count-weighted response is not automatically the same quantity as a simple integral of energy over a rectangular wavelength interval.

If a source emits anisotropically, an isotropic-equivalent luminosity inferred from one viewing direction need not equal its actual total power. If dust or nearby material absorbs and re-emits radiation, a measurement also needs a clear boundary around what is being called the source.

One portion of a spectrum need not represent the whole

Consider an original energy-accounting example with arbitrary power unit P₀. Two stipulated stars each have bolometric luminosity 20 P₀. In one specified hypothetical energy passband, star A emits one tenth of that power and star B emits one quarter.

Their band luminosities are therefore 2 P₀ and 5 P₀. At the same distance, with isotropic emission and a transparent path in that band, B’s received band flux is 2.5 times A’s. Their total electromagnetic powers nevertheless remain equal.

The fractions are supplied teaching assumptions. They are not two computed stellar atmosphere spectra, a claim about a named filter, or a universal relation between temperature and band fraction. The remaining 18 P₀ and 15 P₀ lie outside the specified band in this accounting.

A bolometric estimate from limited observations must account for unmeasured wavelengths. A bolometric correction incorporates a spectrum or empirical calibration and its conventions. It is not justified by labelling the brightest measured band as the total.

An unresolved image can combine different stars

An image that looks like one source may contain an unresolved binary or a chance alignment. Within a defined aperture and calibrated band, incoherent stellar fluxes add. The resulting signal does not automatically belong to a single physical star.

For a second original case, suppose two unresolved components contribute 3 and 1 flux units in the same band and exposure. After the stipulated background correction, the measured combined contribution is 4 units. Assigning all four units to the brighter component overestimates that component’s flux by one unit.

Relative to its true contribution of three units, the overestimate is 1/3, about 33.3 percent. Relative to the measured total, the companion supplies 1/4, or 25 percent. Those percentages answer different questions because their denominators differ.

Higher angular resolution, repeat observations, spectral separation or orbital evidence may help identify a blend. Each brings its own sensitivity and model limits. Subtracting a companion contribution without evidence merely replaces an unknown blend with an assumed one.

Carry definitions and uncertainty into the stellar comparison

A useful brightness comparison names the band or spectral range, calibration system, extraction method, epoch and uncertainty. A useful luminosity estimate also states the distance, extinction treatment and spectral or geometric assumptions connecting the received signal to emitted power.

Some uncertainties affect many measurements together. A shared calibration offset does not disappear when several exposures are averaged, and a precise flux measurement cannot supply an unknown distance. Better precision in one input should not be presented as certainty in the complete inference.

Selection matters when comparing populations. A sample selected by an apparent-flux threshold favours sources that deliver enough light in its selection band. It is not automatically a representative sample of all stellar luminosities, colours or multiplicities.

The next lesson develops the magnitude and colour notation used to report these comparisons. That notation changes how ratios are written; it does not remove the physical distinction between flux, spectrum, distance and luminosity.

Worked examples

Try the reasoning

1. A stipulated source radiates 30 P₀ in total and 0.20 of it in a specified energy band. What is its band luminosity?

   Show the answer and reasoning

   6 P₀.

   Multiply the total power by the supplied fraction: 30 × 0.20 = 6. This does not specify its distance or received flux.

2. Two unresolved components supply 6 and 2 flux units. If all 8 units are assigned to the brighter component, what is its relative flux overestimate?

   Show the answer and reasoning

   2/6 = 1/3, about 33.3 percent.

   The excess is two units and the component’s true contribution is six. The companion’s fraction of the measured total is instead 2/8 = 25 percent.

3. Does equal flux in one calibrated band imply equal bolometric luminosity?

   Show the answer and reasoning

   No.

   Distances, attenuation and the distributions of emitted power across wavelength may differ.

4. Does calling a luminosity intrinsic imply that it cannot change with time?

   Show the answer and reasoning

   No.

   It identifies the power attributed to the source. Variability, epoch and angular emission can still matter.

What to carry forward

• Specify received flux, band luminosity or bolometric luminosity before comparing stellar brightness.
• A measurement in one passband requires spectral information or calibration before it can represent total power.
• Unresolved components, calibration, distance and sample selection remain part of the inference.