PixInsight Forum (historical)
PixInsight => General => Topic started by: mmirot on 2019 October 30 05:43:12
-
What the heck causes this? This is a integration of 195 OIII images. The artifact is not present on calibrated images. In fact I can't see in a stack of 30. It then becomes more and more apparent as the stack increase over 30 images.
-
Do you use a sensor with Anti Blooming Gates? Maybe this artifacts are related to ABGs.
If this is the case, the effect should be related to the exposure time of the subframes. Perhaps try with shorter exposure time and a larger number of frames.
Bernd
-
It is 16803 on SBIG camera. It is ABG sensor. Blooming is in the vertical direction. The image has not been rotated although some were flipped for meridian crossing.
These are 195 10 min exposures. Less time would not appropriate for NB images with this sensor with dark skies(IMO).
I have not seen artifact this posted here or elsewhere.
The raw frames were examined by SBIG and look fine to them.
Thanks
Max
-
OK, it was a guess, but the fact that the artifacts are horizontal precludes this possibility. Sorry, I don't know what could cause this effect.
Bernd
-
Max,
You used Maxim Dl? You also used Maxim's exposure sequence utility?
-adam
-
Yes,
The images are acquired using MaxIM DL.
Thanks
-
Does any one think this could be part of a clamping artifact on low background noise NB images?
I definitely see the related grid artifact too on my OIII images. ( the grid artifact discussed in a separate posting ). https://pixinsight.com/forum/index.php?topic=13205.0
-
Max, you can easily check if this is the case:
Take a close look at the histogram and the Satistics of some calibrated light frames. The question is: are the calibrated light frames clipped at the left side? In HistogramTransformation set the horizontal zoom high enough (perhaps to 100) that this is clearly visible. In Statistics, watch the value of 'count'. Ideally it is 100 %. The deviation from 100 % (= 100 % - count) are clipped pixels.
In this way you can also check whether an output pedestal applied during ImageCalibration (in the section 'Output Files') eliminates the issue.
Bernd
-
Max,
When Maxim DL uses its sequence to acquire images- the chip is read/flushed differently than when you acquire images one at a time say as commanded through another software (like ACP) or even if you pressed the expose button one at a time. It has something to do with the way Maxim handles images with its sequencer. I never got around to telling Doug about this (and never saw anyone complain about it really).
I have some old images (using the same chip) that display this behavior. I can't put my finger on a good one right now.
Here is a subtle example of the same thing (http://www.caelumobservatory.com/mlsc/n7009seq.jpg)
To the right of brighter stars you will see darkness in the horizontal direction. It is always a horizontal effect.
The brighter the star..the easier to see the effect. (this field didn't have any stars really bright enough)
-adam
-
I believe the bias and dark were pure MaxIM DL. The light frames were MaxIM DL commanded via Voyager Automation.
-
Also see this thread: https://www.cloudynights.com/topic/682800-unknown-ccd-artifact/
Bernd
-
The suggestion is this is caused by the Antiblooming gate. The thread questions why it is not vertical like blooming.
Is the ABG a horizontal artifact ??
The only way to fix this to introduce more read noise with shorter exposures ?
Questions no answers. Good to know that I am not the only one to see this artifact.
Thanks
Max
-
https://www.cloudynights.com/topic/435840-banding-anti-blooming-on-alta-f16m-kaf-16803/page-2
Possible clocking problem.
Max
-
How would deal with this in post processing ?
Max