Hi folks. I am imaging with a Altair Astro quad band filter and have had issues with dark subtraction causing negative values and truncating the lights to almost black in PixInsight . See attached. I believe the solution is too add a pedestal offset to the light files. So how do you do this ? Or can it be done in Ekos?
Or if someone outer could tell me is there something else I should be doing to combat by dark files having a higher mean value than my lights.
Last edit: 1 year 1 month ago by Lenny. Reason: Too many entries
You put the pedestal value at the calibration of lights and make sure that the box to subtract this is ticked at integration ( I think) . However I think that it is unusual to need to use pedestals.
I would check that your darks have been integrated correctly. I have found that it is better to integrate darks uncalibrated and as these contain the bias signal there is no need to use a master bias in lights calibration. So try using uncalibrated darks and no bias and see if this makes the difference. I have never found that it is necessary to use a pedestal since I have been using uncalibrated darks.
Just a suggestion.
I am using an Altai Astro 26C camera but shooting with a narrow band Quad filter. So I have read the mean background for these images is less than that of the dark. I have attached the Fits from my Dark/Light for comparison both shot at the same conditions Gain 100 Offset 20.
So I was using the WBPP script in PI, and i did try using the Output pedestal, but the images seem to like detail or contrast. At the moment when shooting, I'm leaving out the darks as I cant seem to get them to work either manually or with WPBB. I have attached the stats to show what I mean.
I am not familiar with the scripts that you are using. Maybe my version of PI is a bit old .
I am always a bit cautious of scripts especially batch pre-processing as if something goes wrong it is difficult to track it down.
I think that the problem is likely related to mismatching between lights and darks (you don't mention if you are using flats as these can introduce problems too). As Wouter suggests things like offset values being different can create the problems that you are seeing.
I think that pedestals can be useful to prevent clipping but clipping often is caused by problems in calibration. I have never found using a pedestal to help much in my image processing.
You may need to give more info. Is this a new problem? e.g. related to a change in your workflow or related to a new piece of equipment e.g. your quad band filter. Or has the camera always been associated with these issues?
The supplied image certainly looks strange. This is a single calibrated light frame, yes? Let us know how you process your darks in terms of calibration, gain, offset and exposure (are these consistent with your light exposures?). Do your uncalibrated light frames show any of these issues?
If the camera is new and this has been a problem from the start check Cloudy Nights or Stargazers Lounge as there may be other users who have encountered these problems and have solutions.
"I am using an Altai Astro 26C camera but shooting with a narrow band Quad filter. So I have read the mean background for these images is less than that of the dark. I have attached the Fits from my Dark/Light for comparison both shot at the same conditions Gain 100 Offset 20."
What you are saying here makes no sense.
The background is not really what you are measuring and for a light image will not be less than the dark (on average they should be identical - that is the point of taking a dark). The filter has no impact whatsoever.
Are both taken at the same temperature?
Losmandy G11 with Gemini 2 controller
QSI 583/683 monochrome camera with filter wheel
Starlight Xpress Lodestar X2 guide camera
Microtouch Focuser controlled using Arduino UNO emulating Feathertouch controller
Raspberry Pi 2GB with Raspberry Pi OS 64bit Buster
It only happens when i use lights with my narrowband filter, I haven't used flats for that very reason until I can narrow down this issue. All my lights and darks are taken in the same session with the same temp, gain and offset.
Yes, this is a calibrated light (stretched )its been clipped due to the delta in the background mean between Light and dark. Maybe I have some light leakage in my darks that I should probably check for that.
Thanks for the response, Yes both are taken at the same temp, offset and gain. On review of my dark images, I may have some light leakage which is causing the higher background values.
Im going to look into that.
You could be right that light leakage may be the problem. You should be able to spot this easily in your master dark by stretching the image.
I usually perform my darks in a more controlled environment and build up a library of master darks for all the usual exposure times that I use. I am consistent with the temperature that I use (-10) and always use the same gain and offset etc. So this makes a library of master darks pretty convenient and allows more time to concentrate on getting enough lights flats and dark flats etc. Using a cooled camera also makes this really easy.
I will offer what I experienced this week in case it is related.
For the first time I captured a few narrowband images (M16) and also for the first time tried running WBPP. Took Bias, Darks and Flats to match the Light setting including temp.
Using ASI1600MM up to 5 minute exposures there was a lot of amp glowin the Darks and Lights. The actual signal was weakest using the Sii filter.
I ran WBPP with all the processing steps through integration checked. It chugged along for a while and then terminated processing the Sii filter with a message along the lines of:
"Select background area with a higher level of signal, it could have high level of clipped data."
One section is on setting a Pedestal value so in WBPP I set the value to 100 and then the WBPP script could run to completion. That is as far as I got yesterday, with 3 master images that sort of seem okay.
That is very interesting. I am using PI 1.8.7 and therefore a earlier version of the WBPP script and have never encountered this info pop up warning. Although I must admit to using batch pre processing rarely in the past and never lately. Lenny, what version are you using?
Thanks for the link to the video guides, they are a great resource!