Bad stars are the most information-dense failure in astrophotography: the SHAPE tells you the cause, if you read it systematically. Three questions sort nearly every case: Which direction do they elongate? Is it the same across the whole frame, or worse in corners? And does it change frame to frame, or sit constant all night?
Uniform elongation, same direction everywhere
Every star stretched the same way across the whole frame is MOTION — something moved during the exposure. If the long axis lies along RA (roughly the direction stars drift with tracking off), suspect tracking/guiding in RA: periodic error unguided, or RA corrections losing the fight. If it lies along Dec, suspect declination problems — backlash oscillation or polar-alignment drift being 'corrected' in lurches. If elongation appears only in SOME frames, stack the good ones and go find the event: wind gusts, a cable pulling taut at a particular mount attitude, or someone walking on a wooden deck.
Elongation that's constant all night in the same screen direction while the guide camera reports round stars is the classic signature of DIFFERENTIAL FLEXURE — guide scope and imaging scope disagreeing about where 'still' is. The cure is stiffness or an off-axis guider; no software setting fixes geometry.
Corner-dependent shapes
Center sharp, corners bad — now it's OPTICS, not motion. Read the corners like a compass. All four corners stretched radially (pointing at the center, like a comet's tail aimed outward): coma or field curvature — your scope needs its flattener/corrector, or the one you have sits at the wrong distance. Corners stretched tangentially or cross-shaped: astigmatism, usually backfocus error again or a pinched optic. One or two corners bad while the others are fine: TILT — the sensor isn't square to the light path; the fix chain is mechanical (check spacer stack seating, camera connection slop, then a tilt plate).
Backfocus deserves its own sentence: most flatteners want 55 mm ±0.5 mm to the sensor. A millimeter off shows in the corners of a modern large sensor. It is the most common optical complaint in the hobby and it costs $20 in spacers to fix.
Rotation and the odd cases
Stars increasingly trailed the farther they sit from one point in the frame — arcs, not lines — is FIELD ROTATION: your polar alignment is off (equatorial) or you're on an alt-az mount without derotation. The rotation center is the guide star if guiding, which is the tell: guiding holds one point, rotation happens around it.
Fat-but-round stars aren't eggs but belong here: focus drift as temperature falls (refocus hourly or on a temperature trigger), dew forming on the corrector (heat straps), or simply worsening seeing. The trend line over the night tells you which.
- Diagnose from the subs, not the stack — stacking blends multiple failure signatures into mush.
- Change ONE thing per night. Fix-everything nights teach you nothing when the stars improve.