Description
Vignetting is a darkening of the edges and corners of the image caused by light loss at the periphery of the field (aperture stops, optical train, undersized corrector). The qualifier "residual" means it persists after calibration.
Normally, flat frames fully correct vignetting by measuring the actual illumination of the sensor. It remains visible when flats are absent, poorly acquired, or mismatched to the optical train as used.
Its signature is radial and symmetric: all four corners darken equally around the optical center, with a smooth transition toward the brighter center.
It is distinguished from a light pollution gradient (oriented in a single direction) and from a mismatched flat (which over-corrects and brightens the corners too much). Dust donuts belong to the same calibration chain.
Visual signature
All four corners of the image are darkened symmetrically, with a brighter center and a smooth transition between the two.
The darkening pattern is circular or elliptical, centered on the optical axis, with no preferred direction (unlike a directional gradient).
The defect appears on stretching: in the linear image it can go unnoticed, but as soon as the shadows are stretched, the center-to-edge luminosity difference becomes obvious.
It is often accompanied by a color shift in the corners and complicates sky background measurement, and therefore color calibration and gradient removal.
Differential diagnosis
Do not confuse with a light pollution gradient: vignetting is radial and symmetric around the center; a gradient is oriented in a single direction (toward the ground or a city).
Distinguish from a mismatched flat: a poorly matched flat over-corrects and makes the corners brighter than the center, whereas uncorrected vignetting leaves them darker.
Separate from dust donuts: these are localized dark rings, whereas vignetting is a global and progressive darkening of the edges.
Common root cause: absent or mismatched flats. A correct flat set corrects both vignetting and dust in one step.
Probable causes
- No flat frames in the calibration workflow
- Flats taken under different conditions from the lights (setup modified)
- Optical train modified (focus, rotation, filter) between lights and flats
- Undersized corrector or reducer, producing strong vignetting
- Large-format sensor exceeding the well-illuminated image circle
- Dew shield or aperture stop partially obstructing the beam
Course of action
- Take flat frames every session, on the exact same setup as the lights
- Do not disassemble or refocus between lights and flats
- Redo flat frames per filter and after any camera rotation
- Expose flats to 30-50% of dynamic range (centered histogram)
- Choose a corrector or reducer whose image circle matches the sensor
- Verify the absence of obstructions (dew shield, aperture stop) in the beam
- Complement with a light gradient removal pass for any residual after flats
The Doc's advice
Vignetting is not a defect, it is an optical inevitability: every optical train vignettes to some degree. The only real answer is clean flat frames taken on the same setup as your lights, without dismantling anything in between. The rule that saves hours: do not touch anything (focus, rotation, filters) between the lights and the flats. Aim for a flat exposure that places the histogram at 30 to 50% of dynamic range, neither clipped nor underexposed. If dark corners persist after calibration, your flats do not match: redo them rather than patching things up with DBE, which would only mask the problem.
Think you can see this defect in your image?
Run a diagnosisFrequently asked questions
Why do my corners remain dark despite using flats?
Because your flats do not exactly match your lights. Vignetting is very sensitive to the geometry of the optical train: if you adjusted focus, rotated the sensor, changed a filter, or dismounted and remounted the camera between lights and flats, the illumination measured by the flat no longer lines up with that of the lights. The solution is to redo flat frames on the exact setup used for the exposures, without touching anything. Also verify that the flats are properly calibrated (dark-flats or bias) and that their exposure level is correct.
How do I take good flat frames?
On the exact same setup as the lights, without disassembling or refocusing. Use a uniform illumination source: a flat panel, twilight sky flats, or a well-lit white screen. Set the exposure for a histogram centered at 30 to 50% of dynamic range, neither saturated nor too dark. Take a set of flats per filter and redo them after any camera rotation. Finally, calibrate the flats with dark-flats (or bias frames) to obtain a clean master flat. Around twenty flat frames per filter is generally sufficient.
Vignetting or gradient: how to tell them apart?
By symmetry. Vignetting is radial and centered on the optical axis: all four corners darken equally. A gradient (light pollution, the Moon) is oriented in a single direction, brighter on one side. The distinction matters because the correction differs: vignetting is corrected by flat frames (calibration), while a gradient is removed in processing (DBE, GraXpert). If a directional brightness slope remains after good flats, it is a gradient, not vignetting.
Can vignetting be corrected without flats, in processing?
Partially, but it is a workaround. A gradient removal tool (DBE, GraXpert) can reduce moderate vignetting by treating it as a low-frequency background component. The result is, however, less clean than with real flat frames: there is a risk of suppressing signal in the corners, skewing photometry and color, especially on extended nebulae that reach the edges. Flat frames remain the correct method because they measure actual illumination; processing only estimates it.