Getting started

Your first session, end to end. Read once, then bookmark for reference.

Install and launch

Starship ships as a Windows installer (Astroworx-Starship-Setup-vXX.YY.Z.exe, Windows 10 and 11, 64-bit). Run it: it installs per-user into %LOCALAPPDATA%\Programs\Astroworx Starship, adds a desktop and Start-menu icon, and can launch Starship at the end. Upgrading is the same installer over the old one; your config.toml, plans and images are kept.

Launch from the icon. Starship starts its web service and opens the dashboard in its own window at http://localhost:8765 (you can also open that address in any browser on the PC, or on the LAN once you bind [web] host to 0.0.0.0).

First launch on a new computer shows the activation screen. Paste the activation key from your supporter email and click Activate; the computer registers with the licence server and the dashboard unlocks. From then on the licence refreshes itself daily and keeps working offline for the paid period plus a grace window - see the Licence & editions topic.

You need on the same PC:

(ASCOM is on by default in a fresh install) - or Alpaca devices reachable on the network;

pointing.

Developers can also run from source: python scripts/run_starship.py in the source tree (Python 3.12, pywin32, numpy, psutil); a source run is never licence-gated.

Connect your equipment

Open Settings. Equipment lives in a per-role tab — Mount, Camera, Focuser, Filter wheel, Rotator, Flat device, Observatory, Power. Open the tab for each role you have hardware for and pick a transport:

Tick Auto-connect on service start if you want that device to connect every launch. Click Save & restart at the bottom — the transports re-initialise on restart.

Don't forget the Site / Location tab: set your latitude, longitude and elevation before any on-sky work. Coordinates left at zero silently break focus-star altitude, airmass and the scheduler.

Once configured, connect and drive everything from the Console (top ⚡ Connect all); the Equipment page is the live status view. It lists every configured slot — your devices show up as tiles right away, offline with their ProgID, and fill with live data the moment they connect. If a device won't connect, the status/error text in the tile or badge is usually clear enough; confirm it works in its own driver / setup tool first.

First exposure

Go to Console. The Camera card has Object, Filter, Exposure (s), Binning, Gain, and Frame (Light/Dark). Enter a short exposure (1 second), pick a filter, click ▶ Capture. The frame lands in your IMAGES directory and appears in the Last captured frame preview on the Console automatically (double-click it to enlarge). A live progress bar tracks the exposure and download; Abort cancels an in-flight capture.

This proves the camera + filter wheel chain works end-to-end. If the frame is upside-down, sideways, or has weird pixel values, fix that before you trust autofocus or the sequencer.

Set the cooling target

If your camera has a cooler, the Cooling section folds into the bottom of the Camera card on the Console (live sensor temp, setpoint, Δ from setpoint, power %, and a state pill).

  1. Settings → Camera → set the cooling Default setpoint (e.g. -10 for a typical CMOS at moderate ambient).
  2. Save & restart.
  3. Console → Camera card → Cooling → set a target °C and click Cool (or Off / Warm up).
  4. Watch the temperature drop. It'll take a few minutes. The Δ turns green and the state reads settled when the temperature stays within the settle window for the dwell period.

Power % at 100% means the cooler is maxed out — you're asking for a setpoint it can't reach; ease the target up. You can leave the cooler on between sessions, or use Settings → Camera → On-connect action ("Turn cooler ON") to start it at the default setpoint whenever the camera connects. Leave it on "Leave cooler in current state" if you'd rather not thermal-cycle the sensor when you reconnect during testing.

Calibrate autofocus

Once per optical setup. Don't skip this — autofocus without a learned sweep is just a focuser-position lookup.

  1. Manually focus first. Use the focuser move controls (type a position then Go, or the −100 / +100 buttons) and Console → Capture to bring HFR/HFD low.
  2. Aim at a normal star field — a good spread of stars in the centre. Tracking on, mount settled. Avoid empty patches and dark nebulae. Broadband filters give the most stars.
  3. Console → Focuser card → run the calibration. Two options:

- Adaptive First Light… (recommended) — self-finding: it measures the seeing/noise, probes the focuser to discover a useful step size, builds the V-curve, then saves the calibration and the Refocus recipe. No step size or range guessing required. - Wizard… — a single adaptive sweep starting from roughly-focused; you can set the step, or leave it blank to reuse what worked last time.

  1. Watch the live V-curve fill in over roughly 5–15 minutes. During the first minute the plot may be empty while Starship measures noise and probes the focuser. Don't touch anything.
  2. When done, the calibration readout shows focus <pos> · step <n>. You're done — for this scope, on these optics.

To test the calibration: move the focuser off, then click Refocus (optionally choosing a filter first). It resolves the apex around the current position in a handful of exposures and reports "focus point found at position X". Go to focus drives the focuser straight back to the last saved position; Clear cal wipes the saved state (you'll need to re-run the calibration).

Build a sequence

Go to Sequencer.

  1. Click New. Give the plan a name and a File id. Pick a Mode: Rotate per frame (cycle filters each frame) or Finish each filter (all of one filter, then the next).
  2. Targets tab → + Add target → enter a name, then fill RA/Dec one of three ways: ⌖ SESAME (resolve the name via CDS Sesame, or from the built-in OpenNGC catalogue when the servers cannot be reached), ⊙ Use current (adopt the mount's current pointing), or ↩ CdC (pull the selected object from Cartes du Ciel). Add exposures with + Add row — each row is a Filter, Frame type, Exposure (s), Count, Bin and Gain.
  3. Constraints tab → set altitude min/max (e.g. 30° minimum), hour-angle limits, or an end-by time. A tripped altitude/HA limit skips that target and moves on rather than aborting the run.
  4. Focus tab → enable refocus triggers: every N frames, every X minutes, on filter change, or on focuser temperature drift. (During a run these fire Refocus, using the recipe you calibrated above; a mid-run refocus failure is journaled but doesn't abort.) You can also set a fallback focuser position.
  5. Cooling tab → tick "Cool sensor before first capture", optionally with a setpoint override and "Wait for settle". If settle is required and cooling fails, the sequence aborts.
  6. Guiding tab → optional PHD2 guiding + dither cadence (only acts when PHD2 control is enabled in Settings).
  7. Meridian tab → for German-equatorial mounts, choose ignore / halt / flip and set how far past the meridian to act, plus recenter-after-flip.
  8. On Start / On Error / On Suspend / On Resume / On End tabs → the actions that wrap the run. On Start opens the roof, unparks, points precisely and autofocuses; On End parks, closes the roof and warms the camera so the rig is always left safe. On Error and On Suspend make the rig safe if something goes wrong or an unsafe event pauses the run.

Validate to check the plan, Save, then ▶ Start sequence. Watch the journal scroll. (You can also Save as template to hand the plan to the multi-night scheduler.)

Things that go wrong

Where to next

For deeper specifics, browse the Modules and Reference sections.

Astroworx Starship v22.2.10 · this page is the in-app help of that buildDownload · HTTP API