The short version

Everything plugs into one box on the telescope, and you run the night from a phone or tablet.

It replaces the laptop, the extension lead, the folding table and the long cable back to the house. Plate solving, autofocus, guiding and the whole imaging sequence are driven from whichever screen you are holding, and shortening the cable runs to a few centimetres on the scope itself removes most of the failures that lose a session. The cost is a closed ecosystem: ZWO cameras and a defined list of mounts, rather than anything that speaks ASCOM.

What it actually is

An ASIAir is a small Linux computer with a handful of USB ports, a 12 volt input, a few switched power outputs and its own Wi-Fi radio. It sits on the telescope, everything else plugs into it, and you control the entire rig from a phone or tablet. That is the whole idea. It replaces the laptop, the extension lead, the folding table and the long trailing cable that used to run back to the house.

Either works, and the app is identical on both. In practice I use a tablet almost every time. The preview frame is what you spend the night looking at, and checking focus, framing and star shape is easier on a larger screen, particularly when the target sits near the edge. The autofocus curve and the guiding graph are both easier to read at a glance too. A phone is still the better thing for a quick look from indoors, or as a spare if the tablet is flat, and you can move between the two mid-session without disturbing anything, because the run lives on the box rather than on the screen.

It matters more than it sounds. Almost every session I have lost was lost to something mundane — a USB cable working loose in the cold, a laptop lid closing, a hub dropping a device mid-run. Shortening the cable runs to a few centimetres on the scope itself removes most of that class of failure at a stroke.

Everything hangs off one box, and the box talks to your phone or tablet phone or tablet Wi-Fi — no cable needed ASIAir control box + storage one 12 V input Mount Main camera Electronic focuser Guide camera Dew heater goto, tracking, meridian flip exposures, cooling, storage autofocus by V-curve guiding and dithering switched 12 V, kept warm all night
The blue lines are short cables that live on the telescope. The only thing running back to the ground is power. On my own rig the controller, main sensor and guide sensor are all inside the ASI585MC Air, so three of those boxes collapse into one.

A night, in the order it happens

The app is arranged around the sequence a session actually follows, which is the reason it is quick to learn even if you have never used capture software before.

How a session runs, start to finish Polar align Plate solve Autofocus Frame on-screen, ~5 min then goto target EAF V-curve rotate, re-solve Autorun Meridian flip Live stack Save & show lights + dithering then darks & flats builds on screen saved as one image Leave it running Watch it build
Only the first four stages need you awake. After framing, the night goes one of two ways. An Autorun runs unattended: it handles the meridian flip itself, then re-solves and re-centres so the framing survives it, and finishes on the calibration frames. Live stacking is the other path, building a single image on screen as each frame arrives.

Polar alignment

The routine takes two exposures a set distance apart in right ascension, plate-solves both, works out where your polar axis actually points and then draws live arrows telling you which way to turn the altitude and azimuth bolts. It does not need Polaris to be visible, which is the single most useful thing about it if you image from a balcony with an obstructed north.

The ASIAir app on a tablet showing a 60 second preview exposure of the Orion Nebula, with the target search, mount slew controls and exposure settings around the edge of the frame
A single 60 second preview of the Orion Nebula, straight out of the camera and onto the screen. This is the frame you use to check focus, framing and exposure before committing to a run, and it arrives about a minute after you ask for the target. The controls sit around the edge so the picture keeps the middle of the screen. This is a tablet, which is what I reach for most nights; the phone layout is the same with less room around the frame.

Plate solving and goto

Plate solving is the feature that changes the hobby. The camera takes a short exposure, the software identifies the star pattern against a catalogue, and the mount then knows precisely where it is pointing rather than where it believes it is pointing. Ask for a target, and it will slew, solve, correct and re-solve until the object is on the pixel you asked for. Star alignment routines, hand controllers and finder scopes all become optional.

Autofocus

With an electronic focuser attached, the ASIAir steps the focuser through a range of positions, measures star size at each, fits a V-curve and returns to the minimum. It takes about a minute and it is more accurate than I am with a Bahtinov mask. More to the point it can be set to repeat automatically every hour or after a given temperature drop, which is what actually keeps a long night sharp — a refractor moves focus measurably as the tube cools.

Autorun, guiding and dithering

Autorun is the plan: a list of exposures, a count, a filter, and optionally several targets in sequence. Guiding runs alongside it, making small corrections from the guide sensor, and dithering nudges the mount a few pixels between frames so that fixed sensor noise lands in a different place each time and averages away during stacking. Dithering is free and it is the cheapest noise reduction available; leave it on.

Live is the other way to spend the same night. Instead of writing frames away for later, the box stacks them as they arrive and shows the result building on the screen, getting cleaner every minute. It is the wrong choice if you want the best possible picture, because you keep the stack rather than the frames that made it, and it is the right choice if there are people standing next to you. On a public night in the desert, a nebula appearing on a tablet in real time does more than any finished image shown a week later.

The ASIAir interface during a session, showing an annotated star field with IC 2177 and two nearby NGC clusters labelled, a histogram along the bottom and a guiding graph reporting 1.51 arcsecond RMS
Three readouts at once, mid-session. The annotation overlay has plate-solved the field and labelled what is in it, here IC 2177 with a pair of nearby clusters. The histogram along the bottom shows where the exposure is sitting, and the guiding graph is reporting 1.51″ RMS total. As a rule of thumb, keep guiding error below roughly half your image scale in arcseconds per pixel and the stars stay round.

Where it is strong, and where it is not

The strength is integration. Everything talks to everything else, nothing needs configuring, and there is no driver layer to go wrong. Frames are written to the onboard storage, so a tablet going flat or wandering out of range does not stop the run — the box carries on regardless, which is exactly what you want at three in the morning.

The cost of that is a closed ecosystem. It expects ZWO cameras, and it supports a defined list of mounts and accessories rather than anything that speaks ASCOM. If you already own a camera from another manufacturer, this is not the controller for you, and a mini PC running NINA will be more flexible and considerably more work. It is a deliberate trade: less choice, far fewer things to debug in the dark.

Four things worth knowing before your first night

  • Station mode is better than the default. Out of the box the ASIAir broadcasts its own network, which means your phone or tablet leaves your home Wi-Fi to connect to it. Joining it to your house network instead lets you keep both, and lets you check on a run from inside.
  • Give it a proper power supply. A single 12 V feed runs the controller, a cooled camera, the mount and the dew heaters. Undersized supplies cause failures that look like software faults, especially when cooling and slewing happen at the same time.
  • Take flats before you take anything down. The moment the camera rotates in the focuser, that night's flats stop matching. It is a two-minute job at the end of the session and it saves a re-shoot.
  • Copy the data off before the card fills. The onboard storage is generous but not infinite, and a night of short subframes fills it faster than expected.

Everything in the recent gallery was captured this way, on the Askar V riding the Juwei 14. If you want to see what the box is actually plugged into, the rig tour walks through every part of it. Once the sequence is running the work moves indoors, which is where Siril takes over.