15th Feb 2024: Astro Pixel Processor 2.0.0-beta29 released - macOS native File Chooser, macOS CMD-Q fixed, read-only Fits on network fixed and other bug fixes
7th December 2023: added payment option Alipay to purchase Astro Pixel Processor from China, Hong Kong, Macau, Taiwan, Korea, Japan and other countries where Alipay is used.
For a couple of months I've been having a real headache with trying to apply flat calibration to my integrations. Results have always been worse than what I get when I don't include flats at all and calibrate only with a MasterDark and a BPM.
I have been thinking that I was doing something wrong, as I have only recently progressed to using a mono astrocam (QHY294M Pro) after previously using a DSLR.
What I have found tonight suggests that there is something wrong with how my MasterFlat is being applied, maybe due to an error in my process for integrating, maybe something wrong with how I'm capturing flats, or maybe even due to an error in APP.
Because my integration results when a MasterFlat was included seemed to exaggerate vignetting and dust spots compared with the results from only MasterDark and BPM calibration, I tried to use an inverted MasterFlat in place of the normal MasterFlat and was amazed to see that the integration result appears to be perfectly calibrated!
This is my Ha integration calibrated with only a MasterDark and BPM:
This is the same set of lights integrated with the same MasterDark and BPM, but with a MasterFlat also used:
This is the MasterFlat on its own:
And finally, this is the integration result using the same MasterDark and BPM as used in the first 2 images, but also using the inverted MasterFlat:
All of the above looks to me like the lights are being multiplied by the MasterFlat rather than divided - obviously I have no idea how this could be happening, and hope that it's simply the case that I'm doing something wrong within APP. I have never experienced this issue before, but maybe there's a difference in how the files from my new camera are handled compared with the files from my DSLR.
I hope that someone can point out what I'm doing wrong, but at least I appear to have found a workaround and can spend the next few cloudy nights reprocessing some images that I've been excluding flat calibration from up until now.
To me, the masterflat doesn't look right. You're expecting a nice and even illumination of the frame and here it looks like you have patches. How do you take the flats, what is your setup exactly and would that explain patches in the flat? The reason that it's not correcting like it should is because the flats are not matching the illumination profile of your lights.
I don't believe the problem is with the illumination pattern of the MasterFlat. If I stretch the lights aggressively I see the same pattern, so that tells me that my MasterFlat is a good representation of the illumination of my sensor with this particular filter. Maybe it's the result of reflections from the filter surface, but whatever the reason for the illumination pattern, it is clear that the MasterFlat calibration is having the opposite effect of what I expect on my integration result, i.e. the brightest areas in the MasterFlat are resulting in increased brightness of the integration when it should be the opposite.
Ok, but they apparently don't represent the exact same, I can have a better look if you can upload say 10 flat frames and 10 lights, together with 10 bias and 10 darks. 10 of each you have basically. 🙂
I'll be happy to upload 10 of each when I get home from work, but it will be Dark Flats and not Bias frames, and I appreciate you volunteering to take a look.
I do think however, that this is missing the point a little. I fully accept that I could be doing something wrong when capturing my data, but that is largely irrelevant to how my results are appearing. My MasterFlat may look odd, but that doesn't explain why the integration result when using it appears brighter in the same areas that the MasterFlat is brighter.
If there is an issue with the data in some way, then that would be relevant I think, that may explain completely why you have the result you show. Again, I can only tell with the actual data.
I have now uploaded 10 each of lights and flats, as well as 10 darks for each exposure length. One thing I noticed during the upload is that I think I may have different offset values for my darks and all other file types. I have been capturing lights and flats with offset=50, believing that to be the same value I used when I created my dark library but on inspection it appears that my darks had offset=60. All other settings (gain and temperature) do match, and I don't know whether the discrepancy in offset could make a difference or not. I hope that's what's causing the issue, because I can easily create a new dark library.
That is very likely the issue indeed, a different offset will change the background noise and APP uses that for proper calibration, I would suggest indeed to try again with new darks.
Thank you Vincent. I plan to shoot a new dark library and will redo my previous integrations with new MasterDarks. I will let you know whether it resolves my issue or not.
I can now confirm that after taking new dark frames with matching offset values to my lights, flats and dark flats, the calibration appears to be working correctly. The image below is just an autostretch of the integration result which included a new masterdark and new masterflat and it seems to have dealt with the uneven illumination and the dust spots. The new masterflat still has the same pattern as the one above, but it's now correcting properly.
