New artifact with new camera

Aaron Linderbaum

Well-known member
I have been using the Altair 269 camera for about 7 months or so and just recently (last 3 weeks & 2 imaging sessions) have noticed a vertical line artifact that I have never noticed before. From what I have read, this is from a camera problem but I am not sure if this is correct.

Below is a link with stacked lights and darks from sessions from 2 consecutive nights (Leo Triplet and M 81). I also included a stacked image of NGC 2264 from a month or so ago.





A few questions: 1) is this artifact a result of a defect in the camera or do all cameras have this issue? I have had a couple of issue with the Altair camera and am wondering if I just have a low quality Altair camera.
2) Why does this line show up in some master darks and not in others? I have used the same master dark on multiple sets of images and it does not seem to be in the final stacked set all the time.
3) Why, when using the same master dark, does the line show up in one stacked set of lights and not in others? Why is it inconsistent?
3) Is this a concern long term?

I guess I am trying to determine if my camera is faulty and should be replaced since it is less than a year old. Or, am I doing something to cause this to present itself on occasion?

Thank you in advance for any insight - I am most grateful!!

Aaron
 
while this type of column defect is relatively common on CCD chips, and can be caused by cosmic ray hits, i'm not sure if the same holds true for CMOS sensors. it's probably worth asking Altair if column defects like this are expected.

as for the variability, i'm not sure if that's because these things fail softly and slowly get worse, or for some other reason. one thing is the IMX269 has quite a number of readout speeds and modes and it might be interesting to vary those and see if the problem goes away. i'm not sure if all those modes are exposed in the Altair driver though.

this article should be helpful in mitigating the problem at least. the two scripts are part of the normal PI distribution, so they should already be installed on your machine.


rob
 
Thank you Rob, for the direction. Great help! Can this be done on the stacked/integrated master dark or is it better to do on the individual images? As you can probably tell, I am new to this hobby and very new to PI...I got it 2 months ago. I do love it....just have to be patient and learn it better.

I am not noticing it on the lights. Is that because it is not there or is it just not noticeable due to the data in the light frame?
 
well if the light and the dark have the same artifacts then they should calibrate out. i understood your message to mean that sometimes the darks have this artifact and sometimes the lights do and so the darks and lights do not match. if the bad columns have appeared at some point in time and then persist into the future, that's totally normal, at least for a CCD with a lot of hours on it. again don't know if this is an acceptable defect in a CMOS sensor.

the linearpattern stuff should be done on the lights. first you integrated unregistered (but calibrated) lights which really makes the linear defects stand out. then you run the detection script on that bogus master light and save the defect description file. after that you run the correction script on the calibrated, unregistered lights, then register the corrected lights and integrate them.

but anyhow maybe you don't need to do this at all if your frames look good after calibration.

rob
 
Rob - Pardon me for having one more clarifying questions on the above link. I ahve read through it and am a bit confused. Does the process start at
The LinearDefectDetection Script and everything above that is just an explanation of what we will be doing with the script?
 
yes that is correct, vicent explains the technique which the script is actually automating. in the end you just need to do the two steps i talked about above.

rob
 
Back
Top