Skip to content
SATLAS

Search SATLAS

Find a facility, a science case, a term or a position on the sky

From detector to catalogue

Every archive takes its data through the same four levels, and the data changes shape on the way: detector readout becomes images, time series or event lists, then light curves, spectra and tables. The names differ from archive to archive (Level 0–3 at Chandra and Euclid, Stages 0–3 at JWST, uncal, rate, cal and i2d in file names); the ladder below is common to all of them.

  1. 1Raw

    Telemetry as recorded, in detector units (ADU, counts, rampsup-the-ramp samplingReading an infrared detector repeatedly during one exposure without resetting it, so charge accumulates as a ramp. Fitting the slope gives a count rate and rejects cosmic-ray hits. or raw channels), with instrument signatures intact. Rarely needed for science and often proprietary.

    • Detector readout
    • Visibilities
  2. 2Calibrated

    Single exposures or data segments with instrument signatures removed (bias, dark, flat fieldcalibrationRemoving instrument signatures and putting data in physical units: subtracting bias and dark current, dividing by the flat field (pixel-to-pixel sensitivity), correcting non-linearity and fitting the astrometric solution. Reference files hold these calibrations and are updated as the instrument is understood., non-linearity, gain) and an astrometricastrometryMeasuring positions and their changes on the sky: parallax, proper motion and orbital motion. Gaia reaches tens of µas for bright stars. or timing solution attached, in physical units.

    Return here when the standard extraction does not suit the science, such as crowded fields, non-standard apertures, faint or extended sources, or a custom treatment of systematics.

    • Image
    • Time series
    • Cube
    • Spectrum
    • Event list
    • Visibilities
  3. 3Measured

    Usual starting point for analysis

    Per-source or per-event quantities extracted from calibrated data: photometryphotometryMeasuring the flux of a source in a passband. Aperture photometry sums pixels inside a region; PSF photometry fits the point-spread function; forced photometry measures at a position fixed in advance, whether or not a source is detected there. and light curveslight curveFlux of one source as a function of time. The basic product of transit searches, variable-star work and transient follow-up., extracted spectra, screened event listsevent listThe basic X-ray and gamma-ray data product: a table with one row per detected photon (arrival time, detector position, sky position, energy and quality flags). Images, spectra and light curves are all binned from it., triggers.

    • Spectrum
    • Catalogue
    • Time series
    • Event list
    • Sky map
    • Cube
    • Posterior samples
  4. 4Combined

    Usual starting point for analysis

    Multi-epoch or multi-field combinations (coaddscoaddMany exposures of the same field resampled onto a common grid and combined, reaching deeper than any single exposure and rejecting artefacts. A mosaic stitches adjacent fields., mosaics, cubesdata cubeA three-dimensional array with two sky axes and one spectral axis (wavelength, frequency or velocity), produced by integral field units and radio interferometers.) and catalogues, often with derived physical parameters such as distance, massarcsecond1/3600 of a degree (″). A milliarcsecond (mas) is 10⁻³″ and a microarcsecond (µas) 10⁻⁶″. Atmospheric seeing limits ground telescopes to about 0.5–1″ without adaptive optics. or redshiftredshiftz = λ_observed/λ_emitted − 1, the stretching of light by cosmic expansion (or by recession). For distant sources it stands in for distance and look-back time: z ≈ 1 is about 8 Gyr ago, z ≈ 10 about 13.3 Gyr..

    • Catalogue
    • Image
    • Cube
    • Sky map

AlertsSearch in real time, from calibrated data

Low-latency notices of new or variable sources, issued seconds to days after detection for follow-up, ahead of full processing.

Each mission's products, level by level

Grouped by the kind of data, because that decides what the levels hold. Under each term is where to get it; underlined terms open the product's entry on the mission page.

Each mission's terms for each processing level, grouped by kind of data
Mission1Raw2Calibrated3Measured4CombinedAlerts
Pixel detectorsCCDs and infrared arrays. Calibrated exposures are images in physical units; photometry, light curves and spectra are extracted from them.
TESSProduct map →
Detector readout
No distinct product
Kepler / K2
Cube
Time series
Catalogue

Kepler Objects of Interest (KOI) and DR25 catalogues

No distinct product
Hubble
Detector readout
Image

Calibrated and CTE-corrected exposures

*_flt.fits, *_flc.fits

Spectrum

Extracted spectra

*_x1d.fits

ImageCatalogue

Drizzled images; Hubble Source Catalog

*_drz.fits, *_drc.fits

No distinct product
JWSTProduct map →No distinct product
Euclid
Detector readout

Level 1 (LE1) raw frames

SpectrumCatalogue
ImageCatalogue
No distinct product
Rubin
Detector readout

Raw exposures

raw

Catalogue
ImageCatalogue

Coadds and the object catalogue, in annual data releases

deep_coadd, object

Alert

Alert stream through community brokers, 60 s latency target

SPHEREx
Detector readout

Level 1 raw images

Image

Level 2 spectral images (Quick Release)

No distinct product
Sky mapCatalogue

Level 3–4 all-sky spectral maps and source catalogues

No distinct product
Catalogue and spectroscopic surveysNo images are released. The products are per-source tables and spectra, reprocessed in full for each data release.
GaiaProduct map →
Detector readout
DESI
Detector readout
Spectrum

Extracted, sky-subtracted spectra per exposure

SpectrumCatalogue
Catalogue

Large-scale-structure catalogues and BAO measurements

No distinct product
Photon countingEach photon is recorded with its time, position and energy, so the core product is an event list; images, spectra and light curves are binned from it.
Chandra
Detector readout

Level 0 telemetry

Event listSpectrum

Level 2 event list; spectra and light curves

evt2

Catalogue
No distinct product
Scanning sky mapsDetectors sweep the sky continuously. Time-ordered data are projected into frequency maps, then separated into astrophysical components.
Planck
Detector readout

Time-ordered data

Time series
Sky mapCatalogue

Component-separated CMB and foreground maps; likelihoods and parameters

No distinct product
InterferometryThe primary data are visibilities, one per antenna pair; images and cubes come from Fourier inversion and deconvolution.
ALMA
Visibilities
Visibilities
Cube
Cube

Archive-wide products and science-ready cubes

No distinct product
Gravitational wavesEach detector produces one calibrated time series, the strain; there are no images. Events are found by matched filtering and characterised by Bayesian inference.
LIGOProduct map →