Description
Dust donuts are dark ring-shaped patches that appear on the image, caused by dust grains resting on a surface near the sensor (sensor window, filter, corrector lens).
Because the dust is outside the focal plane, its shadow is not a sharp point but a fuzzy disk, often annular (darker at the rim) due to the shape of the converging light cone. Hence the "donut" appearance.
Their size depends on the distance to the sensor and the f/ratio: dust on the sensor window gives a small, relatively sharp donut; dust on a more distant filter gives a large, diffuse disk.
This is a calibration defect that is entirely corrected by up-to-date flats, exactly like vignetting. The flats must be taken under the same conditions, otherwise one falls into a flat mismatch.
Visual signature
Circular or ring-shaped dark patches, fuzzy, are scattered across the image, generally darker at their rim than at their center.
Their position is fixed on the sensor: they appear at exactly the same location on every frame (as long as the optical train has not moved), which distinguishes them from real sky objects.
Their size varies: small and relatively sharp for dust on the sensor window; large and diffuse for dust on a filter or a more distant corrector.
They emerge prominently after stretching on a uniform sky background. Without dithering they remain perfectly stationary; with dithering but without flats they leave residual artifacts after stacking.
Differential diagnosis
Do not confuse with residual vignetting: vignetting is a global, gradual darkening of the edges; dust donuts are localized, discrete rings scattered in the field.
Distinct from an internal reflection or halo: those are linked to a bright light source and form around it, whereas a dust donut is independent of stars and fixed on the sensor.
Do not mistake for a real object (faint galaxy, planetary nebula): a donut is perfectly circular, with regular edges, and above all absent from the DSS reference image.
Shared origin with a flat mismatch: if the flats are not taken at the right time, the donuts are not (or poorly) corrected.
Probable causes
- Dust on the sensor window (small, relatively sharp donuts)
- Dust on a filter or corrector (large, diffuse donuts)
- No flats, or flats that do not correct for the dust
- Dust displaced between taking the lights and taking the flats
- Optical train left open or handled in a dusty environment
- Static electricity attracting dust onto optical surfaces
Course of action
- Take fresh flats at every session to erase donuts during calibration
- Do not disassemble the camera between lights and flats
- Clean the sensor window first (blower bulb, optical brush)
- Then clean filters and the corrector if donuts persist
- Avoid dry cotton swabs; prefer a blower and damp optical wipes
- Handle the optical train in a dust-free environment
- Enable dithering to help reject residuals during stacking
The Doc's advice
Dust donuts are a reminder that dust always wins. Do not aim for absolute zero: with up-to-date flats, they disappear at calibration, which is exactly what flats are for. The trap is taking your flats on another night or after dismounting the camera, because a dust particle may have moved and the flat no longer lines up with the donut. So: flats at every session, on the same optical train. If a large donut persists, clean by working up the chain, sensor window first (blower or optical brush), then filters. And ban the dry cotton swab that scratches.
Think you can see this defect in your image?
Run a diagnosisFrequently asked questions
How do I get rid of dust donuts?
The first solution is not cleaning but flats: a set of flats taken at every session, on the same optical train as the lights, erases donuts during calibration because they measure and compensate exactly the shadow of each dust particle. Physical cleaning only comes into play if a donut is very large or very numerous. In that case, work up the optical chain starting from the surface closest to the sensor (sensor window), then filters and the corrector. Flats first, cleaning second.
Why do my flats not correct the donuts?
Because the dust has moved between the lights and the flats, or the setup has changed. A flat only corrects a donut if the dust is at exactly the same position and size at the time of the flat. If you take flats on a different night, after dismounting the camera, or after rotating the sensor, the donuts no longer coincide and the correction fails, potentially creating bright rings (over-correction). The rule: flats at every session, on the exact optical train of the lights, without dismounting anything.
Should I clean the sensor or rely on flats?
Flats are sufficient in the vast majority of cases and should always be your first reflex. Clean only when donuts become too large, too numerous, or so dense that the flat does not correct them perfectly. Cleaning carries a risk (scratches, smears): always start with a blower, then an optical brush, and use damp optical wipes only as a last resort. A dust particle corrected by a flat is preferable to a sensor scratched by a careless cleaning.
How can I tell if a patch is dust or a real object?
Several clues. A dust donut is perfectly circular, with a more pronounced shadow at the rim, and above all fixed on the sensor: it occupies the same position on every frame, regardless of how the sky tracks. A real object (galaxy, nebula) moves with the sky if the framing shifts and appears in the DSS reference. Simple check: compare with the DSS image of the same area (shown by DocStellar); if the patch is not there but is in your flats, it is dust.