Banding in GX100

[dkogan]

Hi, all.

I'm seeing an issue with my GX100, and would like to know if others see it also and what can be done about it.

If I set the camera to a very long exposure time (>1min), and photograph a scene that has thin, vertical-ish features I get VERY pronounced horizontal banding in these features. This is very reproducible and very annoying since it takes the greatness out of great pictures. One example of a scene where this happens is a night sky with star trails. Some of the trails get the banding. Happens in both raw and jpg.

I can post examples if yall want.

Anybody seen this?

 

[dkogan]

Here's an example. Below is a crop from a scene with a starry sky.

Full direct-from-camera image is at
http://www.cds.caltech.edu/~dkogan/banding.jpg

 

[sizifo]

It could be that this is due to the rotation of the earth. I looked up some discussion about star trails and it seems they become noticeable at around 20 seconds at the wide end, and at shorter exposures at the long end.

Hardly a camera fault if this is the case.

 

[dkogan]

The streaking is certainly due to the rotation of the earth, but the black horizontal bands in some of the streaks are a camera fault.

 

[sizifo]

Sorry I missed the point here. See what you mean. No clue.

 

[kai.e.g.]

It's odd that these dark bands are not evident at all outside the stars themselves. Whenever I've seen banding in digital cameras, its usually more noticeable in the darker areas. It's almost like a tiny moire pattern here. The lines don't even seem entirely parallel. Is this crop processed from a RAW file, by the way?

The full-size image you linked to is indeed beautiful, by the way.

It's also possible that an evil mastermind, either local or alien, was speeding up/slowing down the earth's rotation in a rhythmic fashion that night, just for fun. This used to happen all the time in the 1950's.

 

[Lili]

Before I sent my K100D in for sensor issues, hot pixels mainly, I stated seeing black banding like streaks from bright lights in dark areas.
Nothing like this tho

 

[Per Ofverbeck]

Here's an example. Below is a crop from a scene with a starry sky.

Full direct-from-camera image is at
http://www.cds.caltech.edu/~dkogan/banding.jpg

Well, it looks exceptionally regular, indeed, almost like a stroboscope effect. If we reject the hypothesis that all the stars run on AC current as being too unlikely, the periodicity must be in the camera´s electronics instead.

Have you made any attempt at determining the frequency of the banding (easy if you know the xposure time of course); you say it is very reproducible?

Could it be some kind of noise reduction that makes correction at regular intervals during exposure instead of after the exposure is finished? Possibly far-fetched, but the phenomenon looks distinctly non-random.

 

[dkogan]

Thanks for the replies.

It seems that this happens even with stationary objects. Full (but not very interesting) image at
http://www.cds.caltech.edu/~dkogan/bandingnomotion.jpg

Has nobody really seen this before? I'm finding it hard NOT to see this effect with long exposures.