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.
Hi,
Got some time with a new QHY128 before the clouds came to stay. Can't re-take lights right now. But I'm still curious.
Have some odd amp glow or "enhanced vignetting" from the stacking, plus some long arcs one or two pixels wide - with some but not all targets.
So the first thing I'd like to rule out as a current or future cause of problems: The black borders.
I have not seen similar with my other cameras. I would guess that the driver could strip these banners, but it seems QHY didn't have a dedicated Ascom driver for the camera model yet. (it's listed correctly at the download page, but the file is named after another model)
Anyway, when I crop a little to the left, a lot in top, and some to the right, I come out with pretty much exactly the QHY-specified effective XY.
I worry though, that these borders could confuse the processing. I would of course prefer an easy way of batch stripping the frames. (don't think SGP can, at the time of acquisition)
Second, could relatively small variations in sensor temperature cause such issues?
Worst case, 20 subs of the target with temperature between -25 and -27.3°C. (I had a hard time controlling the temperature of the QHY128, until I found that it liked delta 15 or so)
Then again, some sequences were pretty stable in temperature and still showed amp glow or strange vignetting.
Plus, I used bias frames, and flats, that I acquired using a flat panel.
The darks were taken at -30°C, constantly. So they differ from the lights, and I fear that I have confused the dark scaling, which I guess is active as I can't find such a setting in the config page.
From darks back to arcs. I can mostly find them in subs. Like cutting 60 to 90° pieces of a huge circle made up of thin line just one or two pixels wide, putting them to the right or left in the frame and making them bulge to the center. I was surprised to find in the atlas, that there were no bright stars at or beyond the frame border, radiating these artifacts.
SGP/PHD reported dithering, so I was likewise surprised to find that the very thin arcs were not canceled. But I guess I will have to try different integration settings.
This is not really me. I usually toss away anything without near perfect calibration frames.
Never seen anything like, since I stopped using DSLR.
But, being far from experienced, I guess I've messed up at some point.
Thanks for reading, please share any ideas!
That's a lot of information, thanks. 🙂
So to try to answer them;
- Black borders in the subs might cause problems if the APP processes take this in as signal (noise in that case). I'm not sure if it does as I think it is intelligent enough to not take total black values as something it should take along. When you stack them you might be able to see if the statistics still make a bit of sense for noise and background levels (maybe compare them with previous data from another camera). But having them does indicate a wrong read-out of the camera, so that's not optimal to begin with. Hopefully they get the correct driver up soon!
- Tiny temperature differences shouldn't be a major problem, I doubt a 2 oC makes real differences in noise levels.
- Dark scaling can be checked in the Calibrate tab (2) where you select the options for a masterdark. Having those at -30 should be fine, again it's like a 5 oC difference. It might be a real difference at that point, but I again doubt if that's big.
- The arcs and such I can't really judge, can you send a few examples?
If these black borders could cause problems, would it then make sense
to use for example an image container and do a dynamic crop on all of the
lights before feeding them into APP???
Just a thought,
Mert
Thanks for helping.
(sorry about the late response)
The driver settings had a little bug, was eventually able to exclude the overscan area with the black borders. New data should be fine.
I asked the camera vendor about the "arc-artifacts" (added a picture to the thread https://www.qhyccd.com/bbs/index.php?topic=6833.0), but I guess the solution is in dealing with some odd internal reflection that I have not yet been able to find. Will try a different scope.
Anyway, it is now clear that the latter has nothing to do with APP.
Didn't nail the dark scaling setting at first, better now!
