How to plan an imaging session

Two clear hours on the forecast rarely become two hours of light. The camera downloads after every sub and the guider settles after every dither. Change filter and the focuser hunts. By dawn the stack is shorter than the night looked.

Most of a good imaging session is decided in the afternoon. You fit the target, the sky and the rig together, and you are honest about the pauses. The guides cover the weather. This one covers the plan.

Deciding whether tonight is worth a session

It starts with darkness, and with cloud where the target will be rather than straight overhead.

Then weigh the stretch against the effort. A full setup, with polar alignment and focus, needs a long dark stretch to be worth it. A rig that lives in an observatory, already aligned, can use an hour and a half.

Twilight is worth less than it looks. The sky is brighter, and the brighter background swallows faint detail.

If a second model keeps the sky clear for longer, treat those hours as a bonus rather than building the plan on them. The clear sky forecast guide and the weather model guide explain what to compare.

Picking a target that's properly up

The best hours for any target are around transit. It is at its highest, and its light crosses the least air.

That matters most for blue and OIII, which lose more light low down than Hα and SII. If the night includes an OIII or blue run, give it the hours around transit and let the red filters take the lower ones.

Work from west to east. Catch whatever is setting before it sinks into the murk, and leave rising targets until they have climbed. An altitude curve makes the order obvious, and the Targets page draws one for every object, with the dark band, the Moon and the flip marked.

On an equatorial mount, find out early where the flip falls. A target that transits mid-session will break the run whether you planned for it or not.

Tracing your own skyline

The horizon that matters is the one from where the tripod stands. It is usually a tree, a roof or the neighbour's extension.

Low obstructions often cost nothing. If a tree, plus your safety margin, stays under your minimum altitude, you were never going to image through it anyway. The tall ones on the target's path are the problem, and they tend to announce themselves when a galaxy slides behind the chimney at one in the morning.

It is worth tracing the skyline once, from the spot you actually use. The horizon tool does it with your phone's camera, and needs no account. Move the tripod a few metres and trace it again.

Framing before the night

Catalogue sizes describe the bright part of an object. The faint outskirts are often what makes the picture, so frame against those.

Rotation matters too. A few degrees can be the difference between a comfortable fit and clipped spiral arms, and trying it first in the Sky Atlas, with your own camera and focal length, saves doing it in the dark.

Mosaics need overlap to stitch, and overlap costs sky. Each panel covers less new ground, so a two-panel job can tip into three. Every panel also needs its own integration, and the hours multiply before you count the slews between them.

Filters and sub length

Let the target choose the filters. Emission nebulae glow in Hα, OIII and SII, so HOO or SHO suit them. Galaxies, clusters and reflection nebulae shine by starlight and want broadband or LRGB. A bright Moon pushes emission targets towards narrowband and makes galaxies a harder sell.

Narrowband subs run longer for a simple reason. A narrow filter passes very little sky, so it takes minutes for the background to bury the camera's read noise. Broadband under a bright suburban sky gets there far sooner.

Set a total for the whole project and expect it to take several sessions. Then let the first sub of the night settle the exposure. You want the histogram's peak clear of the left edge, few clipped stars, and guiding and focus that hold. The seeing and transparency guide covers how much detail a given night is likely to offer.

What a session actually collects

This is where most plans come unstuck. Dividing the clear hours by the sub length gives a number the night never delivers, because a session is full of small pauses.

Every sub has to download. Every dither needs the guider to settle before the next exposure. A filter change means the wheel turning and usually a refocus, and a new target or mosaic panel means a slew, a plate solve and a recentre. Focus also drifts as the temperature falls. On an equatorial mount, the flip costs a few minutes while the mount swings over and the guider finds its star again. Only whole subs count.

Here is what that does to a two-hour block, with 2 s downloads and a 15 s settle after each dither. With 60 s broadband subs, dithering every third sub, it holds 107 subs rather than the 120 you get from dividing time by exposure length. Narrowband is usually dithered every sub, and with 300 s subs the block holds 22 rather than 24. Split 3 h between two narrowband filters: 90 min, a 2 min filter change including refocus, then 88 min. At 300 s per sub that gives 17 + 16 = 33 subs, or 2 h 45 min of light, rather than 36. Those examples start on target, so moves, the flip and any extra refocusing come on top.

Plan my night does these sums for you once you have an account, a saved site and a saved rig. It counts whole subs with your own timings for downloads, dithers, filter changes and moves, and on an equatorial mount it ends blocks at the flip. If cloud wrecks the plan, “Replan remaining night” starts again from now, and the subs you log or import count towards your projects.

When the cloud comes in halfway

It helps to have chosen a fallback before dusk, such as a short run on the filter furthest behind.

On a broken night, pick one filter and stay on it. Changing filters in every gap spends the clear sky on the wheel and the focuser. When a bank passes, check a fresh sub before trusting the run, because thin cloud often trails the thick stuff. The cloud cover guide has more on working the gaps, and the forecast and nowcast guide covers the evening check.

At the end of the night, bank the subs worth keeping and record each one once, so nothing is counted twice. Then give the next clear spell to the filter with the fewest chances left this week.