Diagnosing Baked-In Concentric Rings

Introductory Comments

Many consumer cameras (DSLR and mirrorless) perform undocumented raw-data processing which is impossible to disable. Raw-data filtering to reduce image noise is one example of this and is often destructive to stars. Certain other kinds of internal raw-processing lead to coloured concentric rings or waves to appear in the background of images.

Although these rings are rarely seen in a single exposure, heavy processing can reveal them. Unfortunately the typical workflow for astrophotography, designed to reveal faint ojects in the image by stacking, background subtraction and intensity stretching, will also reveal any faint rings caused by the camera's own internal processing.

This kind of internal raw-processing is often colloquially referred to as "cooking" the data and the resulting rings are "baked-in". They can be baked into both the night-time exposures of the scene being imaged and the flat-frames used for calibration.

Causes of Concentric Rings

Rings in astro-images can be caused for a variety of reasons. The design of the optics, internal reflections, stray light and dew on optical surfaces can all potentially lead to rings and they are not related to the camera's internal processing. Faults in the astro-processing workflow can also be a problem e.g. high-order polynomial background extraction or processing in a low bit-depth integer format instead of floating point. Where rings are caused by the camera's internal raw-processing there are still a number of distinct causes:

• Concentric rings caused by digital scaling, for instance on the Sony A7S
• Concentric rings caused by Nikon's lossy compression
• Concentric rings caused by Nikon's image correction, possibly a sensor colour-shading correction
• Concentric rings caused yet-to-be-diagnosed raw-processing in many FujiFilm cameras and some Canon cameras

A list of camera models with known problems can be found in this camera summary

An excellent example of the kind of rings being discussed is this one from the Nikon D5500, used with the kind permission of Michael Covington, author of "Digital SLR Astrophotography", in his blog:

    
Click on image for larger version

Test Protocol

The purpose of the test protocol is to distinguish the camera's internal raw-processing from other potential causes.

Here is an outline of the test protocol that will produce good data for revealing the rings, if they exist. A sequence of 7 low-ISO raw flat frames at different exposure-levels is required, acquired as follows:

• Where the options exist, switch the camera to lossless compression or uncompressed with 14-bit output (not 12-bit)
• Switch off in-camera lens corrections
• Use ISO 100
• If using a lens (rather than a telescope), do not allow the aperture (i.e. f-ratio) to change during the imaging sequence
• Shoot the first flat using a "normal" photography exposure level with the peak of the back-of-camera histogram slightly to the right of centre.
• For each of the 6 subsequent flats, reduce the exposure by half between each shot by doubling the shutter speed. Obviously the final shot will be extremely underexposed but that's the intention.
• The analysis of each exposure will subtract the bias, debayer (in floating point format, not integer), bin the data (to reduce noise) and then divide the red channel by the green channel before intensity stretching. Afterwards, divide the blue channel by the green channel and stretch.

If the analysis of each exposure shows rings (or wavy steps) and the number of rings (or steps) changes from exposure to exposure then this is clear evidence of some kind of internal raw-processing.

Example Results

Here is a typical result showing 6 exposures from the sequence (for a Canon EOS R8), dividing the red channel by the green:

    
Click on image for larger version

Notice the progression of rings from one exposure to the next.

Another interesting result is this one from a Nikon D5100:

    
Click on image for larger version

Three exposures from the sequence are displayed. The top row shows the red channel divided by the green and the bottom row shows the blue channel divided by the green. Notice the progression of stepped waves in the top row and the progression of concentric rings in the bottom row. Concentric rings are not always central in the image - it depends on the internal processing that caused them.

Conclusion

Many consumer cameras show evidence of internal raw processing leading to coloured rings or waves in images. As awareness of the problem grows within the astro-imaging community, the issue is being identified in an increasing number of cameras.

If your camera does not appear in my summary of cameras with known artefacts but you think your camera may be suffering from the problem then I'm always happy to analyse a sequence of exposures taken using the above protocol.

Last updated by Mark Shelley: 8 October 2023 (astro@markshelley.co.uk)