Processing the Andromeda Galaxy: Why the Core Isn't the Problem

Search for advice on processing a bright galaxy core and most of it is really advice about nebulae - protect the highlights, don’t let the centre blow out. Applied to the Andromeda Galaxy, it solves a problem M31 mostly doesn’t have.

Andromeda’s core is compact and bright, but it is nowhere near the dynamic-range problem an emission nebula like Orion presents - the nucleus does not dominate the frame’s brightness range the way a nebula’s ionised core does. The actual difficulty with M31 is almost the opposite: the vast majority of its light comes from a disk so faint it sits barely above the sky background, professional surveys of it work down around magnitude 28.5 per square arcsecond, several magnitudes fainter than the sky most of us are shooting under to begin with. Losing that disk is not a stretch problem. It is a background-modelling problem.

The real risk: subtracting the galaxy along with the gradient

Every frame carries a light-pollution gradient, and removing it means fitting a smooth surface to what the software believes is sky and subtracting that surface out. Andromeda’s outer disk is faint enough, and extensive enough across the frame, that a background-extraction pass with samples placed carelessly will read genuine spiral structure as “sky” and fit a surface straight through it.

The result looks clean rather than damaged, which is what makes this particular mistake so easy to miss. You get a tidy background and a galaxy that quietly ends at a smaller radius than it actually does, its faintest outer arms gone along with the light pollution.

This is exactly the failure mode a low-order, sky-only gradient fit exists to prevent - keep the fitted surface’s freedom low enough that it cannot bend to follow real structure, and keep its samples confined to genuine sky rather than anywhere merely dim. For a target this extended, that discipline matters more than almost any other single processing decision.

Framing before you process anything

M31 is roughly three degrees across including its full outer halo - about six times the width of the full Moon - which is wider than most telescope and camera combinations can capture in one frame at any focal length that also resolves real detail in the disk.

Decide before you shoot whether you are framing the bright core and inner disk at real resolution, or the whole galaxy including its faint outer extent at a wider, softer scale. Trying to do both in one frame with the wrong focal length is the most common reason a first M31 attempt disappoints - not the processing afterward, but a frame that was never going to hold both the target and the room around it.

If you do want the whole structure at full resolution, this is a mosaic target: multiple overlapping panels, each stacked normally, then blended by their shared stars into one wide field. That is a different pipeline stage from stacking a single field, and it is worth planning for before the first exposure rather than discovering afterward that one panel does not cover what you needed.

The companions are part of the frame

M32 and M110, Andromeda’s two brightest satellite galaxies, sit close enough to the main disk that most standard framings include at least one of them, and a slightly wider frame gets both. They are worth keeping deliberately in shot rather than centring M31 so tightly that a genuinely interesting part of the system gets cropped out - three galaxies for very little extra effort is a better frame than one galaxy alone.

Star colour, carefully

A galaxy field is a star field first and a galaxy second by pixel count, and Andromeda’s disk sits in front of and among a real population of foreground Milky Way stars along with the resolved blue supergiants and star-forming regions in its own spiral arms. Photometric colour calibration, set from a star catalogue rather than assumed from the frame’s average colour, is what keeps a foreground star reading as the temperature it actually is instead of picking up a colour cast from the galaxy’s own light nearby. Skipping this step is how a processed M31 photo ends up with every star in the frame reading a slightly wrong, uniformly-tinted colour - technically stars, but not quite believable ones.

The short version

• Andromeda’s problem is a faint, extended disk near the sky background - not a blown-out core. Treat it as a gradient-removal problem, not a highlight-protection one.
• Keep the background fit low-order and its samples confined to real sky, or you will subtract genuine outer structure along with the light pollution.
• Decide your framing - core and inner disk, or the whole ~3° extent as a mosaic - before you shoot, not while processing.
• Frame M32 and M110 in deliberately. They are close enough that most shots include one anyway.
• Calibrate star colour from a catalogue so the foreground starfield reads true rather than tinted by the galaxy beside it.