Back to forums

Artefacts in Photo

Sony A7III Artifacts Help Needed
12 replies•400 views
Armin Lukas avatar
Armin Lukas avatar
Hi,

I am looking for an explanation of some artefacts in a photo. It is only found in the light frames, and still there after calibration of frames.
It might be caused by the moon or other light sources, but not sure. Have you ever seen such a thing?

Here is an example of it:






A few nights before there was not such a thing.

Can this be removed in post processing?

Scope: Sharpstar 13028HNT
There was a dew cap on it.
Shun-Chia yang avatar
Shun-Chia yang avatar
If the camera model, shooting parameters such as gain offset exposure, the software used for shooting, and the file format are provided, it would be very helpful.
Armin Lukas avatar
Armin Lukas avatar
Shun-Chia yang:
If the camera model, shooting parameters such as gain offset exposure, the software used for shooting, and the file format are provided, it would be very helpful.

Sure:
Sony A7 III, modified (with new UV/IR Cut Filter)
ISO100
240 sec. exposure
364mm- F/2.8

Run by ASIAIR Plus
Dithering on
Brian Puhl avatar
Brian Puhl avatar
Are you using pix to stack?  Not sure I've seen this one before, but just shotgunning here, did you make sure linear defect correction is unchecked?

All this is assuming you have proper bias frames as well, darks can be skipped since you dithered, and cosmetic correction can be added in its place.
Armin Lukas avatar
Armin Lukas avatar
Brian Puhl:
Are you using pix to stack?  Not sure I've seen this one before, but just shotgunning here, did you make sure linear defect correction is unchecked?

All this is assuming you have proper bias frames as well, darks can be skipped since you dithered, and cosmetic correction can be added in its place.

Yes, but also happening when stacking with SIRIL. It makes no difference if linear defect correction is on or off.
I noticed it gets better with flats and looks much worse without the bias.  My theory is it must be something wrong with flats.
Brian Puhl avatar
Brian Puhl avatar
Armin Lukas:
Brian Puhl:
Are you using pix to stack?  Not sure I've seen this one before, but just shotgunning here, did you make sure linear defect correction is unchecked?

All this is assuming you have proper bias frames as well, darks can be skipped since you dithered, and cosmetic correction can be added in its place.

Yes, but also happening when stacking with SIRIL. It makes no difference if linear defect correction is on or off.
I noticed it gets better with flats and looks much worse without the bias.  My theory is it must be something wrong with flats.



Try shooting longer bias frames.    At least 0.2 second or better..    You can always look at your master flats and bias stretched in Pixinsight, if theres issues, you should be able to notice them.   I wouldn't expect that to be a flat issue.    I presume you aimed for 50% histo/adu for your flats correct?
Helpful Concise
Armin Lukas avatar
Armin Lukas avatar
Brian Puhl:
Armin Lukas:
Brian Puhl:
Are you using pix to stack?  Not sure I've seen this one before, but just shotgunning here, did you make sure linear defect correction is unchecked?

All this is assuming you have proper bias frames as well, darks can be skipped since you dithered, and cosmetic correction can be added in its place.

Yes, but also happening when stacking with SIRIL. It makes no difference if linear defect correction is on or off.
I noticed it gets better with flats and looks much worse without the bias.  My theory is it must be something wrong with flats.



Try shooting longer bias frames.    At least 0.2 second or better..    You can always look at your master flats and bias stretched in Pixinsight, if theres issues, you should be able to notice them.   I wouldn't expect that to be a flat issue.    I presume you aimed for 50% histo/adu for your flats correct?

I tried longer bias, tried dark flats: with no success. There is only a problem in the light frames, so it must be any kind of light source nearby. Otherwise I can't find an explanation.
I have never hat this kind of problem with this camera.

Will do some tests at another location and will report.
Jean-David Gadina avatar
Jean-David Gadina avatar
How did you take your flats? Using some LED panel?

If so it can be a refresh rate issue.

Can you show how it looks like on a single flat?
Dave Rust avatar
Dave Rust avatar
Hi, Armin!

This looks to me like nothing more thanthe passage of high cirrus clouds. Try shooting again when you know the sky is clear. Does the artifact again appear? Also, the colors are similar to that of a solar aurora. As you know, they've been active lately. Either way, I suspect the artifact is not your gear, it's something external.
Helpful Concise
V avatar
V avatar
This is a common issue on the A7III's, including my own. You have a specifically bad case of it.

It is mostly a readout error.

The only courses of action I can tell you is to take 100+ bias frames.

Worst part about this issue is that it also affects the 410 chipped astrocams.
Armin Lukas avatar
Armin Lukas avatar
This is a common issue on the A7III's, including my own. You have a specifically bad case of it.

It is mostly a readout error.

The only courses of action I can tell you is to take 100+ bias frames.

Worst part about this issue is that it also affects the 410 chipped astrocams.

Hey thanks for replying! Could you share some more info about it? Other forums links?
I’ll try more bias and see if that works.
Armin Lukas avatar
Armin Lukas avatar
Well well, I found a (good) enough solution to this problem.

First af all, there is definitely a problem with the sensor. The thing is, this kind of full frame sensor is stitched from two parts.
This is causing the left/right split. The purple/green artifacts are somehow caused by the electronics.

Here is more info about it:
https://www.cloudynights.com/topic/519941-sony-a7s-split-sensor-and-banding/page-4

Let's see what  can be done:

This is the before (stacked and processed with high stretch +calibration with biases and flats) 40x240 sec. ISO100:


You can clearly see the left/right split and the coloured stripes.

First thing to do is to use a ISO of 640 or higher AND expose at least to the middle. This improves the SNR ratio and removes the horizontal stripes (at least it reduces it).
This photo is a stack of a 10x120sec ISO 640 no flats darks and biases.
The coloured lines are not visible anymore.


Next step is to use flats and biases. I exposed the flats the same like the biases: 1/125 of a second ISO 640 (histogram peak at the middle).
The vertical line are removed by the flats. Somehow this didn't work before (probably messed up the flats).

This is a picture after calibrated with flats and biases:


The artifacts are now greatly reduced!
I haven't found any other method, but this seems to work (more testing is needed).
Helpful Insightful Engaging
Tambs575 avatar
Tambs575 avatar
I realize this is an old thread, but I didn’t see where anyone mentioned the banding that can occur with the A7iii’s silent shutter mode activated.  This can happen with the silent shutter turned on when shooting under artificial light, including star light.  Not sure if that is what’s happening here, but I’ve experienced it on my own A7iii.
Well Written Concise