Amp Glow Not Calibrating Out

[Buzzard75]

This is kind of a weird one. I know how to do calibration frames, take my darks, flats, dark flats, etc. I've been doing this for a few years and I've never seen anything like it. I recently purchased a full set of Chroma 31mm filters in Ha, SII, OIII, and LRGB. I have successfully taken images with all of them and all of them have been properly calibrated except for my Ha filter. No matter what I do, I always have a little bit of "amp glow" left on the side of my images. Any thoughts on why that might be and how to correct it? It doesn't make any sense to me at all. Thanks in advance!

EDIT:

For what it's worth, I have a ASI183MM camera, which is notorious for "amp glow". I'm using APT to capture all my images including my calibration frames. I process everything in PixInsight manually. I have not used the WBPP script in quite some time. I have gone back and examined some other files that I have done recently with other filters. It appears that it is actually a problem with all of my images, but it is just more pronounced in the Ha. Possibly due to the higher signal to noise ratio. It seems that PixInsight is not doing a great job completely eliminating the dark noise from my frames even though I am using all the same exposure settings. I have read other posts on other forms expressing the same issue. Some have claimed that SkyX does a significantly better job. Any thoughts or tips on something I may be missing in PixInsight that solved a similar issue for others would be appreciated.

Edited December 6, 2021 by Buzzard75

[scitmon]

I had this problem starting out with my ASI 294MC, the problem turned out to be something with the ZWO native drivers I was using.  I was advised to change to the ASCOM drivers which solved the problem.  

[vlaiv]

Two things need to be checked:

1. what is your offset and what is your calibration method (which calibration frames do you take and how you calibrate)?

2. If you are using APT - are you using any sort of "helpers" / "assistants" - like flats assistant that "automatically" do things for you?

[Grifflin]

I have had a similar issue with my 294mc and processing everything in PI.  Today I use  Flats, Dark Flats and Darks for my calibration (so no Bias frames) and, in PI, I turn OFF Dark Optimisation.  Cleared up the problem for me.

Neil

[Buzzard75]

2 hours ago, vlaiv said:

Two things need to be checked:

1. what is your offset and what is your calibration method (which calibration frames do you take and how you calibrate)?

2. If you are using APT - are you using any sort of "helpers" / "assistants" - like flats assistant that "automatically" do things for you?

Using offset 10 for all of my light and calibration frames. I used SharpCap and examined the histogram to determine what the best offset was. A setting of 10 pushes it off the left side of the histogram enough to not clip the black, but not too much. I'm using darks, flats, and dark flats for calibration. I create a dark flat master using ImageIntegration (no Normalization, Weights don't care, Windsorized, and no Normalization for pixel rejection). I calibrate the flat frames with the dark flat master using the ImageCalibration process (unchecking the optimize button for the darks and then create a master flat with ImageIntegration (Multiplicative, Weights don't care, percentile clipping and equalize fluxes for pixel rejection). I then create a master dark with ImageIntegration (same settings as dark flat). I use both masters to calibrate my light frames (Local Normalization, Weight FITS Keyword, Windsorized, Local Normalization for pixel rejection) and then create the stacked images. I'm essentially following the Light Vortex Astronomy tutorials and have been for quite some time with no issues until recently. I attempted to calibrate a dark frame with a dark master just to see what it would do and it also left a bit of glow behind, which I found extremely interesting and very confusing.

I am using APT. I use the flats aid to determine the length of exposure for a specified ADU value and then use that exposure length for both my flats and my dark flats. The broadband filters are usually 1-2 second exposures. I up the brightness of my flat panel a bit and get exposures of 3-4 seconds for my narrowband filters.

 

I do appreciate any insights you may have.

[vlaiv]

All of that sort of looks ok - except all those settings for PI.

Could you possibly upload following for inspection:

- two dark subs (choose one randomly from the beginning of the batch and one from the end of the batch)

- master dark

(this is needed as you say that dark / dark calibration does not remove amp glow completely)

- one flat dark

- one flat

- one light sub

I'm going to inspect all of those and see what I can find. I think we should look into individual subs rather than stack to see if they are working as they should prior to moving onto creating masters.

[Buzzard75]

2 hours ago, vlaiv said:

All of that sort of looks ok - except all those settings for PI.

Could you possibly upload following for inspection:

- two dark subs (choose one randomly from the beginning of the batch and one from the end of the batch)

- master dark

(this is needed as you say that dark / dark calibration does not remove amp glow completely)

- one flat dark

- one flat

- one light sub

I'm going to inspect all of those and see what I can find. I think we should look into individual subs rather than stack to see if they are working as they should prior to moving onto creating masters.

Appreciate the offer of assistance. I'll try and upload some files after work today.

I did do some more research last night and found a discussion about the use of Local Normalization. I started playing around with the Normalization in the image integration and pixel rejection options of Image Integration process. I found that using Local Normalization produced the results mentioned above of not removing the amp glow in the stacked image. When I used Adaptive Normalization though, it was gone. It may be that if I want to continue using Local Normalization that I need to tweak my settings a bit rather than running with defaults. I've tried a few different scales (128, 256, 512). Of those, 512 gave the best results, but still not as good as Adaptive Normalization. I have not had the opportunity to play with the sliders at all to see what effect they would have.

[vlaiv]

31 minutes ago, Buzzard75 said:

Appreciate the offer of assistance. I'll try and upload some files after work today.

I did do some more research last night and found a discussion about the use of Local Normalization. I started playing around with the Normalization in the image integration and pixel rejection options of Image Integration process. I found that using Local Normalization produced the results mentioned above of not removing the amp glow in the stacked image. When I used Adaptive Normalization though, it was gone. It may be that if I want to continue using Local Normalization that I need to tweak my settings a bit rather than running with defaults. I've tried a few different scales (128, 256, 512). Of those, 512 gave the best results, but still not as good as Adaptive Normalization. I have not had the opportunity to play with the sliders at all to see what effect they would have.

Simplest thing you can do to see if it is due to way you integrate - is just use simplest average integration without any normalization / pixel rejections / whatever for everything.

Create master calibration frames and integrate image like that - no weights no fancy stuff - just simple average.

Sure image might not look the best and there will be hot pixels and all - but at this stage you don't care about it  - you just want to see if your calibration works and if using advanced stuff is causing you issues. If you get clean result - then you can start narrowing down for the cause of issue (and be happy as your camera / calibration works as it should).