Methodology

What the archive can claim

A number without its method is only an assertion. Every distance in the deep-sky index carries the technique that produced it, every survey band carries what it actually measures, and every propagated orbit carries how far it can be trusted. This page is that ledger, written out.

Last updated 27 September 2026.

In short

The deep-sky index holds 101 objects. 97 carry a distance, and each one names the method it came from.
Those 97 distances are measured by seven different techniques, from direct parallax on nearby stars to surface-brightness fluctuation on galaxies fifty million light years out.
The atlas draws 21 survey bands across six wavelength groups — optical, infrared, radio, ultraviolet and high energy. Each band states what it is sensitive to, not just its wavelength.
Nothing is fetched at runtime. Every figure on this site is computed in your browser from files in this repository.
Where a claim would be false, the site says so instead. A day-old orbital element set is shown as a day old, and beyond a month the Live Tracker stops predicting passes rather than guess.

How a distance is measured

There is no single way to measure how far away something is. Each technique works over a limited range and leans on a different physical assumption, so a deep-sky index that reports one number per object is quietly hiding that the number means different things in different rows. This index does not hide it: the method travels with the value, and you can read it on any object's card.

The table below is the actual distribution across the 97 measured objects, with a real example from the catalogue for each method.

Method	Objects	Example	What it actually relies on
Parallax	29	M42 · 1,344 ly	The apparent shift of an object against its background as Earth moves. The most direct method there is — it assumes nothing about what the object is, only that the shift can be resolved. Runs out of precision within a few thousand light years.
RR Lyrae variables	30	M13 · 22,200 ly	Old, metal-poor pulsating stars that share a narrow range of luminosity, so their apparent brightness gives the distance. The workhorse for globular clusters — and the most-used method in this catalogue for exactly that reason.
Cepheid variables	15	M31 · 2,537,000 ly	A period–luminosity relation: brighter Cepheids pulse more slowly. Calibrated on nearby parallax distances, then pushed outward. The method that first established that other galaxies lie outside the Milky Way.
Surface-brightness fluctuation	15	M87 · 53,500,000 ly	The graininess of a galaxy's light: nearer galaxies show lumpier individual stars, more distant ones blur smooth. Reaches tens of millions of light years, where individual stars can no longer be resolved.
Spectroscopic estimate	6	M16 · 7,000 ly	Spectral features compared against stars of known type. Used here for nebulae and clusters where the individual stars are hard to isolate against glowing gas.
White-dwarf fitting	1	M4 · 7,200 ly	The fading sequence of white dwarfs in a cluster is matched to a known cooling curve. A precise but demanding technique, used here on a single object.
Expansion parallax	1	M45 · 6,523 ly	For an expanding shell, the angular speed of growth compared with its measured line-of-sight velocity gives the distance — geometry instead of a brightness ladder. Recorded here for the Pleiades.

Why the mix matters. A catalogue that reported only the parallax distances would cover the nearest handful of objects and nothing else. One that reported only Cepheid distances would reach the galaxies and skip the clusters inside our own. The seven methods overlap and cross-check each other, which is the point: where independent techniques agree on the same object, the distance is better than any single one of them. Each distance on this site shows which technique produced it so you can judge the figure on its own terms.

What each survey band sees

A wavelength is not a colour, it is a question. Change the band and a different population of objects answers: young hot stars in the ultraviolet, old ones in the near-infrared, dust and cold gas in the far-infrared, and the non-thermal emission of jets and remnants at radio wavelengths. Switching bands on the Deep Sky atlas is not a filter over one picture — it is asking the sky something else.

Optical

440 nm Blue plate, DSS2 — Blue-sensitive photographic plate, the classic survey look.
660 nm Red plate, DSS2 — Red-sensitive plate, sensitive to H II regions and late-type stars.
800 nm Infrared plate, DSS2 — Far-red plate, reaches through the dust that hides star formation.
620 nm SDSS r, SDSS — Charge-coupled imaging, deeper than the plates but not all-sky.

Infrared

1.25 µm 2MASS J — Near-infrared, dominated by the light of cool giant stars.
1.65 µm 2MASS H — Near-infrared window, straddling the H-band opacity minimum.
2.16 µm 2MASS Ks — Longest 2MASS window, the least affected by interstellar dust.
3.4 µm WISE W1 — Stellar photospheres and the bulk of the old stellar population.
4.6 µm WISE W2 — Tracers of warm interstellar gas and embedded young stars.
12 µm WISE W3 — Polycyclic aromatic emission, the glow of the dust in a nebula.
22 µm WISE W4 — Warm dust close to the stars heating it.
60 µm IRAS — The first all-sky far-infrared view of cool dust.
100 µm IRAS — Cold dust at large scale, the backbone of cirrus maps.

Radio

21 cm 1.4 GHz, NVSS — Continuum radio: jets, supernova remnants and active nuclei.
4 m 74 MHz, VLSS — Low-frequency survey, steep-spectrum emission only.
92 cm 325 MHz, WENSS — Northern low-frequency survey, wide but shallow.
6 cm 4.85 GHz, GB6 — High-frequency survey, biased towards flat-spectrum cores.

Ultraviolet

227 nm Near UV, GALEX — Hot young stars on the upper main sequence.
153 nm Far UV, GALEX — The youngest stellar populations, before dust reprocesses the light.

High energy

5 nm 0.25 keV, ROSAT — Soft X-ray emission: accreting binaries and hot gas.

Composite

visible Colour composite, DSS2 — Infrared, red and blue plates stacked as the red, green and blue channels. The familiar full-colour view is itself a constructed image, and this is the construction.

Why a stale orbit is never called live

The Live Tracker propagates satellites from a dated element set. Each set carries the epoch at which its orbit was fitted, and the accuracy of a propagation decays as the elements age. A site that showed a fresh position without saying how old the underlying orbit was would be accurate-looking and wrong.

The module always shows the epoch, the age and the resulting along-track uncertainty, and refuses to describe a day-old element set as "live". Positions themselves are recomputed every second, which is what the interface means by live. Beyond a month in either direction the module stops predicting passes altogether: a confident time from an element set that stale would be a lie dressed as a feature.

The uncertainty figure is a deliberately conservative order-of-magnitude guide rather than a formal covariance, and it is labelled as such in the interface. It grows with the absolute distance from the epoch, in both directions — an instant pinned before the epoch is just as far from the fitted orbit as one pinned after it, so a one-sided figure would quietly promise accuracy for a date in the past that the elements cannot give.

Where the material comes from

Mission records are drawn from the NASA, CNSA, ESA, ISRO and JAXA public archives. Sky and survey data are compiled from the surveys named in the band list above. The licences of every bundled piece of software, and the sources of every dataset, are recorded on the notice page.

Astral Archive is an independent project and is not affiliated with any space agency.

Open the Sky Atlas Read the notice page Back home