How can I correct radial brightness falloff using flat-field correction?

Asked 1/31/2019

5 views

2 answers

0

I have photos with obvious radial darkening toward the edges. In my case it is not lens vignetting; it is caused by using a small on-camera light source with the subject and background close to the camera.

I’d like an automatic or semi-automatic way to even out the brightness, ideally on Linux. I looked at GIMP and RawTherapee, and I saw that RawTherapee offers flat-field correction, but I do not have a matching reference file.

Can flat-field correction be used for this kind of brightness falloff, and if so, what kind of reference image should I capture?

Originally by Photography Stack Exchange contributor. Source · Licensed CC BY-SA 4.0

Photography Stack Exchange contributor

7y ago

2 Answers

2

Flat-field correction is designed to solve this problem. If you have access to the same camera and lens, you can capture the reference image yourself. Use a plain white wall that is evenly illuminated and slightly defocused, not a picture of the same background and lighting as the image you want to process.

Originally by user75526. Source · Licensed CC BY-SA 4.0

user75526

7y ago

0

AI Answer

Generated from our catalog & community — verify before relying on it.

Yes. Flat-field correction is the appropriate method for this kind of radial brightness falloff, even if the cause is lighting rather than the lens itself.

To use it, capture a reference “flat field” with the same camera/lens setup. Photograph a plain white surface that is evenly illuminated and slightly out of focus. The goal is to record only the illumination pattern and falloff, not scene detail.

You do not need a photo of the same background with the subject removed. Instead, you need a uniform reference frame made under the same setup so the software can measure and compensate for the uneven brightness across the image.

If your editor supports flat-field correction, that reference image can then be used to normalize the brightness in your real photos.

UniqueBot

AI

7y ago

Your Answer