Using ARGOS — a complete session¶

This guide covers one full observing session, from the settings you fill in indoors to the folder you hand to Siril at the end. It assumes ARGOS 0.4.1 and a Seestar telescope.

0.4.1 is a field-validation release

Operate the telescope attended: there is no weather safety and no restart recovery. Keep every raw FITS frame, and do not treat an ARGOS curve as publishable without independent reduction. What the live curve does and does not mean is set out in Differential photometry — method and equations.

The night in seven moves¶

1 · Indoors

Site, observer code, telescope profile, sessions folder. Resolve tonight’s targets while you still have internet.

2 · Seestar app

Power up, run the GoTo pointing calibration, focus. ARGOS does not replace this step.

3 · Connect

Telescope, Camera, Filter Wheel and Focuser must all read Ready.

4 · Plan

Resolve the target, check its altitude, build the steps, name the FITS target.

5 · Solve

Identify the field, then click stars to assign Target, Comparison and Check.

6 · Watch

Read the comparisons as well as the target. A shift shared by everything is weather.

7 · Hand off

Copy the session folder. Reduce it in Siril and star_var_script.

1. The five screens¶

Everything below happens in one of five places, chosen from the sidebar on the left. Learn these now and the rest of the guide has somewhere to hang.

Screen	What it is for
Connection	A two-minute setup stop: find the Seestar, connect the four devices.
Observe	Where the night happens. The image, the camera controls, focus, the mount, the overlays and the live curve are all here, as movable docks.
Plan	Build and run the acquisition sequence. Its heading reads Sequencer; it is the same screen.
Review	Post-session, offline. Curves, quality trends, frames, AAVSO export.
Settings	Observer, site, paths, telescope profile, catalogues.

Two menus matter more than they look:

Field → solving and the catalogue overlays.
Photometry → Photometry setup… (choose the target, review the ensemble) and Live curves & diagnostics… (the curve window).

Prerequisites

This guide assumes ARGOS and ASTAP are installed and linked — if not, do Install ARGOS first; it takes twenty minutes and it is not a field task.

If a term goes past you, Glossary defines every one this documentation uses: AUID, check star, ensemble, zero point, green plane, TG, HFD and the rest.

2. Before you go out¶

Do this indoors, with internet. Catalogue lookups are cached, so what you resolve now still works in the field with no connection.

Settings → Observatory. Enter the observer name and, for AAVSO submissions, the observer code — left unset, every AAVSO export is stamped XXX. Search for your site, pick the right result, correct the altitude if you have a measured one, and save it as a favourite.

Settings → Equipment · camera. Select the telescope you actually own. The S30 Pro is the reference profile; S30 and S50 are selectable but flagged unvalidated — do not use them for precision photometry until their parameters have been validated in the field.

Settings → Files · application. Choose the sessions working folder. FITS frames, the session record and the measurements are written there.

Settings → Catalogues · data. Check the bundled essential catalogue and the CDS / NASA / AAVSO cache paths. Refresh clears only the selected cache; the next explicit search fetches it again if internet is available. Set the sequence presets folder and the camera calibration folder (local PTC files) if you use them — they are distinct from the catalogue caches and ARGOS never modifies them on its own. Enable JSONL diagnostic logging only for a run you intend to debug; those files stay in the session folder and are never transmitted.

Plan ahead. While you still have internet, search tonight’s targets and resolve the fields you will need.

Before you close the laptop

Observer name and AAVSO observer code entered
Site saved as a favourite, altitude corrected
Telescope profile matches the telescope you own
Sessions working folder chosen, with room for the night
Tonight’s targets resolved, so their coordinates are in the cache
ASTAP executable and database both confirmed in Settings → Astrometry

3. Start the telescope in the Seestar app¶

Every night begins in the official Seestar app: power and initialise the telescope, then complete its GoTo pointing calibration before connecting ARGOS.

Do the initial focus there too. ARGOS’s focus tools work and can refocus during the session, but its autofocus is currently slower. ARGOS does not yet replace the official app’s startup procedure or its pointing calibration.

4. Connect¶

Select the telescope in Telescope · equipment.
Enter the Seestar’s IP address and Alpaca port — usually 32323, and 10.0.0.1 in Seestar access-point mode — or press Discover, which searches for the telescope and remembers the last address.
Press Connect equipment. Telescope, Camera, Filter Wheel and Focuser must all read Ready.
Show connection and device details exposes the individual device controls and the Stellarium server if you need to diagnose one of them.

Networks, tethering and offline use are covered in Field connectivity guide.

5. The order of the night¶

This is the sequence, and it is not obvious from the interface. Read it once before your first session.

Connection      connect · four devices Ready
     │
Observe         Shot          → one test frame, check focus and exposure
     │          Field → Identify field   → the frame is solved
     │          Photometry → Photometry setup…  → click the target
     │                                          → accept or pick comparisons
     │                                          → add a check star
     │
Plan            build the steps · Start sequence
     │
Observe         Photometry → Live curves & diagnostics…  → watch it run
     │
Review          open the session · read the quality trends · export

Two points worth stating plainly, because getting them wrong costs a night:

Solve and assign roles before you press Start. The photometry roles are what makes a sequence produce a light curve. Frames taken before a target exists are saved as FITS — nothing is lost — but they produce no measurements, and they are not re-measured when you assign roles later.

The frame you solve on comes from the Shot button. In the Acquisition dock on the Observe screen, set a type of Light, an exposure and a gain, and press Shot. That is a single test frame: it is what you focus on, what you check the exposure level on, and what you solve. The same button provides the pilot frame that automatic comparison selection measures candidates on.

6. Plan the sequence¶

The Plan workspace deliberately separates search, plan parameters, altitude, presets and run controls into movable panels; the step table stays in the centre.

The Scientific source panel accepts designations. Messier, NGC and IC resolve immediately from the bundled catalogue, tolerating lower case, optional spaces and common aliases. HD and other names go through CDS Sesame once, then cache locally.
The result fills the plan’s name and coordinates. It does not slew: verify the target, then command the GoTo from the Observe telescope panel or point with Stellarium.
Target visibility plots altitude across the coming local night, with the dashed line at 30°. It is computed for the site in Settings.
Build the steps: frame type (Light / Dark / Flat / Bias), filter, exposure, gain, count and interval. The estimated duration is an estimate — leave margin for downloads and autofocus.
Acquisition options holds repeat, autofocus cadence and end action. The Target name (FITS) field is the scientific identifier written to OBJECT, into the filenames and into the photometry target set — it is not the session name. A source search fills it; for a manual Stellarium pointing, enter it yourself or accept a catalogue suggestion after checking it. Presets saves and reloads a plan as JSON.
Start sequence begins the run. Pause finishes the current frame then holds; Stop ends the run.

Tip

Keep the acquisition settings stable through a time-series run. Changing exposure or gain mid-run changes the instrumental system you are measuring in.

What to actually put in those boxes¶

The step table asks for numbers and does not suggest any. Here is a defensible starting point — a starting point, not a recommendation for your target. Take a Shot and check it before committing the night.

Setting	Start with	Why
Filter	IR	The IR-cut position is the broad one. `LP` narrows the passband, which changes what your instrumental green is; never mix the two in one run, and do not use `LP` for photometry unless you know why you are.
Exposure	10–20 s	Long enough to be sky-limited rather than read-noise-limited; short enough that a mag-9 star does not saturate and that field rotation does not smear. ARGOS’s own default is 10 s.
Gain	80	The shipped default. Higher gain buys read noise, not signal, and costs dynamic range.
Count	enough to cover the event	For a variable: at least a full cycle for a short-period star, or two to three hours of a longer one. Below ~30 frames the run-level scatter estimate does not exist and your error bars are photon-only.
Interval	0	Back-to-back. Download already costs several seconds per frame.

Check the peak, not the picture

The one thing to verify on your test Shot: click your target and read the peak ADU on its star card. Too low and you are measuring noise; near the top of the range and you are clipping the very star you came for. Adjust exposure first, gain second.

Bear in mind that ARGOS’s own saturation flag is not currently trustworthy — the threshold is not reconciled with the sensor’s full well (Differential photometry — method and equations §3.5). Your eye on the peak value is the better guard.

How long is enough?

Scintillation alone puts roughly 15 mmag on a single 20-second sub through a 30 mm aperture at airmass 1.5, no matter how bright the star (Differential photometry — method and equations §9). If the amplitude you are chasing is 0.3 mag, one frame sees it. If it is 0.02 mag, you need to average tens of frames and you need the night to cooperate. Decide which case you are in before you drive somewhere.

Calibration frames — take them, ARGOS will not remind you¶

ARGOS applies no calibration. But the session folder has darks/, flats/ and biases/ because Siril will need them, and the only chance to take them is while you are still out there.

The filter wheel makes two of the three easy: its first position is a physical Dark slot, so darks and biases need no cap and no cover.

Frame	How	How many
Darks	Filter `Dark`, same exposure and same gain as your lights, same temperature — so take them at the end of the run, not at home	15–20
Biases	Filter `Dark`, shortest exposure the camera allows, same gain	20–30
Flats	Needs an even light source over the objective: twilight sky, or a panel. The telescope is sealed, so there is no other way in	15–20, at half saturation

Note

Flats are the awkward one on a Seestar, and if you skip them Siril will still calibrate with darks and biases alone. That is worth doing: darks remove the hot pixels that Differential photometry — method and equations §3.5 describes faking a dip. Flats remove the vignetting gradient — which matters most when your comparison stars are far from your target, and is exactly why ARGOS keeps the ensemble within 25 arcmin in the first place.

7. Observe¶

Observe is the session workspace. Acquisition, Telescope, Focusing, Image display, Image statistics, Focus diagnostics (HFD/FWHM), Activity log and Differential photometry are docks: drag them by their title bar, stack them, or detach them onto a second screen. The filter wheel has no dock of its own — the filter is a selector inside Acquisition, and in the sequence rows. View → Reset Window Layout restores the default arrangement at next start.

File → Open FITS image… opens an existing frame for inspection and field identification.
Field → Identify field solves the current image; the ASTAP options and the catalogue controls live under the same Field menu.
In the Telescope panel, command the GoTo only after checking the name, the coordinates and the intended target.
The context bar under the image shows the frame name and session progress; detailed statistics stay in their own panel.

Important

Display is not data. Debayer mode, stretch and channel are preview settings. The FITS written to disk stays linear 16-bit CFA with BAYERPAT='GRBG'.

8. Choose the photometry roles¶

After a field identification, click a star in the image. What the card shows depends on what the click matched: for a catalogue object, its identity, type, magnitude range, period, AUID and RA/Dec; for a source matching no catalogue entry, the measured quantities instead — FWHM, HFD, SNR and peak ADU. Drag its title to move it; resize it from its bottom-right corner.

Button	Scientific meaning
Target	The variable or transit host whose differential magnitude is measured.
Comparison star	A reference star that builds the ensemble zero point.
Check star	A star expected to be constant, measured against the ensemble to expose its drift.
Remove	Takes this star out of the saved set.
Dismiss	Closes the card only; roles already assigned are kept.

Check star does not check the focus — it assigns a scientific role. The role buttons stay disabled until the field is solved and the star has reliable coordinates: a wrong WCS gives a wrong identity.

The recommended flow is: solve the field, choose a target, let ARGOS propose comparisons or pick them explicitly, then add a check star.

How many comparison stars? Aim for five.

Three different numbers appear in the settings and they mean three different things:


2 (`min_comparisons`)	Below this a measurement is annotated, not hidden. It is a floor, not a target.
3	Fewer than three stable comparisons and the ensemble reports not ready in the Comparison quality panel.
5 (`auto_comparisons`)	What the automatic proposal will add for you, and the right thing to aim for.

Two comparisons is a bad ensemble: the outlier rejection of Differential photometry — method and equations §4.2 needs three to work at all, and with one the error bar is meaningless. Take the five ARGOS offers unless you have a reason not to.

One more thing the interface will not tell you: match their colour to your target. The zero point differences a catalogue V magnitude against an instrumental green one, so a colour mismatch drifts through the night looking like variability. VSP publishes B and V for every sequence star — keep the ensemble within about 0.3 mag of the target in \(B-V\).

A measurement with fewer comparisons than the floor is flagged as fragile rather than hidden. How the ensemble is built, clipped and propagated into an uncertainty is described in Differential photometry — method and equations.

Selecting stars in a crowded field, catalogue depth and what an unmatched source means are covered in Identify a solved field. For a transit, see Exoplanet transit preparation.

9. Read the live curve¶

The Field photometry window — Photometry → Live curves & diagnostics… — contains:

Light curve — target and check star on the scientific plot; comparisons on their own diagnostic plot, labelled Δmag from own median. Show uncertainties toggles the error bars without changing the data.
Variable stars — the variables identified in the field.
Metrics — airmass, FWHM, sky background, HFD, star count and sensor temperature.
Targets and Comparison stars — the set currently in use.
Aperture · references — the measurement geometry and the VSP references.
Export measurements… — a CSV of target, check and comparisons that preserves the roles and can be reopened in Review.
Export target (AAVSO)… — the scientific target only.

Warning

Comparison curves are diagnostics. Never submit one as an AAVSO observation.

Reading the curve honestly

The error bars include a run-level systematic floor measured from the curve itself, so they are not photon-noise-only optimism — but the measurement is still on uncalibrated sub-exposures in an instrumental band. A dip you can see is worth reducing properly; a dip at the level of the scatter is not a detection. The method, and its limits, are in Differential photometry — method and equations.

A shift common to the target and the comparisons means cloud, tracking, focus or saturation — not variability. That is what the comparison diagnostic plot is for.

10. What ends up on disk¶

A sequence is organised for Siril under the configured sessions folder:

<sessions>/
└── <timestamp>_<OBJECT>/
    ├── lights/
    ├── darks/
    ├── flats/
    ├── biases/
    ├── session.json
    ├── photometry_quality.json
    └── diagnostics/

photometry_quality.json is the leave-one-out assessment of the comparison ensemble described in Differential photometry — method and equations.

Live measurements and target sets are also kept under <sessions>/targets/. When you report an observation, keep the raw FITS, session.json, the CSV exports and any *_diagnostics.jsonl together.

Copy the whole session folder before reducing it in Siril and star_var_script — the folder layout is already the one Siril’s sequence discovery expects, so a copy is all it takes. The naming rules and the full header list are in FITS headers reference.

11. Review a finished session¶

Select Review in the sidebar, then Open session…, and choose the folder that contains session.json — not the lights/ subfolder.
Check the frame count, the filters and the warnings, then the FWHM, HFD, sky-background and temperature trends. These describe the ARGOS session; they do not replace a scientific reduction. Review’s docks are Source light curve, Frame quality — FWHM, Comparison stars, Comparison quality, Frames, Session metadata and AAVSO export.
Comparison quality reports the latest leave-one-out assessment: one row per comparison with its sample count, scatter, median formal error and a live status of Stable, Noisy or Unstable. Select a row to show its curve. A star that never leaves Waiting for 10 points simply has too few usable measurements to judge.
The curves shown are the differential previews recorded during acquisition. Hover a point for its values; select it and press Open source frame to open that FITS in the viewer. The Frames tab gives the same access by double-click, with astrometry and full inspection available there. Nothing is written back to the raw frames.

Review works with no telescope connected. Set the AAVSO observer code in Settings before exporting; XXX means it is missing.

12. Can I submit this to the AAVSO?¶

The documentation warns you repeatedly not to treat a live curve as a publishable measurement, and then offers an Export target (AAVSO)… button. Both are right, and here is how they fit together.

What the export is. An AAVSO Extended File Format file: your observer code, one row per point, the differential magnitude and its uncertainty, JD_UTC as the time (which is what AAVSO asks for), the band TG, CNAME=ENSEMBLE, and TRANS=NO — an explicit declaration that the magnitudes are untransformed. Only curves whose role is target are written: a comparison or check curve can never leave through this door as an observation.

What it is not. Calibrated. There is no dark, no flat, no bias and no colour transformation, the zero point is cross-band, and the uncertainties do not contain slow systematics. See Differential photometry — method and equations §10 for the full list.

The honest answer

Do not submit the live export as your observation of record. Treat it as what it is: a field check that the night worked, a preview to see whether the star did anything, and a well-formed template of your own metadata.

The submission comes from the reduction — calibrate and register the raw frames you kept, re-do the photometry there, and submit that. The raw frames are the deliverable of an ARGOS night; the curve is the reason you know the night was worth reducing.

If you do submit an ARGOS-derived measurement, describe how it was obtained: uncalibrated sub-exposures, untransformed TG against V sequence magnitudes. ARGOS writes na into the NOTES column, so you have to add that yourself before uploading. An observation with an honest note is useful. An uncalibrated observation presented as a calibrated one is worse than none.

Practical submission is through the AAVSO’s own WebObs upload; the file format is documented under Extended File Format. Neither is part of ARGOS.

13. Reporting a problem¶

Start with More → Create local support bundle…. The ZIP is created locally, at your request. It contains redacted technical logs and, if you choose, the session’s JSON/CSV/JSONL metadata. It never includes raw FITS frames, the observer’s identity, coordinates or network address, and ARGOS never transmits it. Check its contents before sharing it.

A useful report names the exact version, the platform, what you did, what you expected and what happened — with the log excerpt around the failure.