Data format
What arrives when you obtain 7DT data: two image products and a catalog, a filename that tells you what a file is, and headers that carry everything needed to turn counts into calibrated magnitudes.
What a data product is
Images
A calibrated single exposure and a coadd are the two primary image products. Coadds may be nightly stacks, commonly around 300 seconds, or longer multi-epoch stacks.
| Product | Filename | Format and size | Contents |
|---|---|---|---|
| Calibrated single exposure | *_100s.fits | FITS image, 32-bit float, 9576 × 6388 px, ~234 MB, image in HDU 0 | Bias, dark and flat corrected single exposure. Pixel values remain in instrumental ADU. |
| Coadd | *_coadd.fits | FITS image, 32-bit float, standard grid 10200 × 6800 px, ~265 MB, image in HDU 0 | Resampled stack on a common grid. Pixel values are flux-scaled to approximately µJy per pixel. |
What a filename tells you
A basename identifies the observation completely, so a file can be placed without opening it. A typical single exposure:
T08147_m650_7DT02_20251126_043413_100s.fits
| T08147 | RIS tile number, or a free target |
|---|---|
| m650 | Filter |
| 7DT02 | Telescope unit |
| 20251126_043413 | UTC date and time |
| 100s | Exposure time |
The same stem carries through to everything derived from that image, distinguished by a suffix.
| _coadd | Coadded image. May be a nightly stack or a longer multi-epoch one. |
|---|---|
| _cat | Source catalog for the preceding image product. Empty primary HDU 0, source binary table in HDU 1. |
| _weight | Coadd inverse-variance weight map, in the primary image HDU. |
| _footprint | Per-pixel contributing-frame information, where produced. Availability depends on the processing route. |
Conventions
| WCS | Single-frame distortion is represented with TPV (TAN--TPV), not SIP. Coadds are resampled onto a TAN grid. |
|---|---|
| Single-frame units | Instrumental detector values in ADU after basic calibration. |
| Coadd units | Flux-scaled to µJy per pixel. |
| Pixel scale | 0.505 arcsec per pixel, so a single exposure covers about 1.34° × 0.90°. |
| Filter identifier | Medium bands are prefixed m, from m400 to m875, with some wider bands suffixed w as in m466w. Broad bands are u, g, r, i and z. |
| Dates | The observing-night date and the UTC date encoded in a basename may differ by one calendar day. |
FITS header keywords
Beyond the standard keywords, a 7DT header carries what the observation was, how good it turned out, what is needed to calibrate it, and where every input came from. Quality metrics are written by the pipeline for every image it produces and are also ingested into the observation database.
What the observation was
| OBJECT | Target or survey-tile identifier. |
|---|---|
| FILTER | Filter used for the observation. |
| TELESCOP | 7DT telescope unit identifier. |
| EXPTIME | Exposure time. For a coadd, the sum of the contributing exposure times. |
| DATE-OBS | UTC observation timestamp. For a coadd, the mean time of the contributing frames. |
| MJD | Observation time as Modified Julian Date. For a coadd, the mean of the inputs. |
| AIRMASS | Airmass of the observation. |
| MOONSEP | Angular separation from the Moon, degrees. |
| IMAGEID | Unique image identifier. |
| PIPE_VER | Py7DT pipeline version stamp. |
How good it is
| RSEP_Q2 | Astrometric quality: representative separation of sources from Gaia DR3, arcsec. Variants RMS, MIN, MAX, Q1, Q2, Q3, P95, P99. |
|---|---|
| ISEP_Q2 | Astrometric quality: representative separation within 7DT images taken together, arcsec. Same set of variants. |
| BIN0FWHM | PSF size as it varies across the field. BIN0, BIN1, BIN2 run from center to corner, equally spaced in squared radial distance. |
| SEEING, PEEING | PSF FWHM, corners excluded, in arcsec and in pixels. |
| ELLIP, ELONG | PSF ellipticity and elongation, corners excluded. |
| SKYVAL, SKYSIG | Median and sigma of the sky background. |
| SATURATE | Pixel saturation level. Read per image, since it can differ between products. |
| SANITY | Pipeline quality control. False marks a frame that should normally be excluded. |
| REJ_PROC | Processing stage at which the image was rejected, where a rejection was recorded. |
How to turn counts into magnitudes
| APER, APER_1–APER_5 | Fixed aperture definitions used for photometry. The SExtractor AUTO aperture is not among them. |
|---|---|
| ZP_AUTO, ZP_0–ZP_5 | AB magnitude zero points for AUTO and for each fixed aperture. |
| EZP_* | Uncertainty on the corresponding zero point. |
| UL5_*, UL3_* | 5σ and 3σ limiting magnitudes for the corresponding aperture. Not defined for AUTO. |
| EGAIN | Effective gain in e⁻ per ADU. Not the same as the camera keyword GAIN. |
Where it came from
| PPFLAG | Bitwise-OR calibration provenance: 0 same-date raw calibration, 1 nearby-date calibration, 2 manually generated master, 4 a SANITY=False input was used, 8 masters matched by ignoring lenient keys, 16 masters matched by ignoring temporal bounds, 32 masters matched by ignoring hard keys. |
|---|---|
| IMCMB###, IMCID### | Input master frame identifiers, recorded in single-image headers. |
| IMG#####, IID##### | Input image identifiers, recorded in coadd headers. |
Two keywords are worth reading before any analysis. SANITY is the pipeline's own verdict on the frame, and False means it should normally be excluded. PPFLAG records every compromise made in finding calibration masters, so a frame calibrated against something less than ideal always says so.
Source catalogs
The naming rule
Py7DT source catalogs are FITS binary tables following SExtractor conventions, holding instrumental measurements, calibrated photometry, source morphology and reference-catalog cross-matches. One rule explains most of the column names: an unsuffixed photometric column is an instrumental measurement, and a column suffixed with a filter name is the calibrated quantity in that filter.
| MAG_APER | Instrumental aperture magnitude. |
|---|---|
| MAG_APER_m650 | Calibrated aperture magnitude in m650. |
| FLUX_APER | Instrumental aperture flux. |
| FLUX_APER_m650 | Calibrated aperture flux in m650. |
The suffix convention extends to the corresponding errors and related quantities wherever filter-specific columns are provided.
Position and shape
| ALPHA_J2000, DELTA_J2000 | Source sky position, J2000. |
|---|---|
| X_IMAGE, Y_IMAGE | Source position in image pixel coordinates. |
| FWHM_IMAGE | Source FWHM in pixels. |
| ELLIPTICITY | Source ellipticity. |
| CLASS_STAR | SExtractor stellarity, for star/galaxy separation. |
| FLAGS | Source-extraction quality flags. |
Instrumental measurements
| MAG_AUTO | Instrumental SExtractor AUTO magnitude. |
|---|---|
| MAG_APER–MAG_APER_5 | Instrumental fixed-aperture magnitudes. Apertures are defined in the image header. |
| MAGERR_* | Uncertainty on the corresponding instrumental magnitude. |
| FLUX_AUTO, FLUX_APER* | Instrumental flux measurements. |
| FLUXERR_* | Uncertainty on the corresponding instrumental flux. |
| SNR_* | Signal-to-noise ratio for the corresponding measurement. |
Calibrated photometry
| MAG_*_<filter> | Calibrated magnitude in the observed 7DT filter. |
|---|---|
| MAGERR_*_<filter> | Uncertainty on the calibrated magnitude. |
| FLUX_*_<filter> | Calibrated flux in the observed 7DT filter. |
| FLUXERR_*_<filter> | Uncertainty on the calibrated flux. |
Gaia cross-match
| source_id | Identifier of the matched Gaia reference source. |
|---|---|
| ra, dec | Gaia reference-source position. |
| phot_g_mean_mag, bp_rp | Gaia broad-band photometry. |
| separation | Angular separation between the detected source and its Gaia match. |
| mag_m400 … mag_m875, mag_u–mag_z | Gaia-XP synthetic reference magnitudes across the 7DS filter set, wide bands included. |
Working with it
The pipeline that produces these files, and the packages for reading and analyzing them, are described under software. How the data are obtained is on data access. Measured depths, zero-point accuracy and delivered image quality — the numbers behind the header keywords above — are on the performance page.