Hi my name is Klaus and I joint this Forum yesterday to get some insight into astronomical image processing.
I have installed AA5 to process my pictures. Over the weekend I took some test pictures to learn how to use AA5.
The camera is a Nikon D800 with a 200mm f2.0 Telelens used at f2.8. I took 5 Lights, 5 Darks, 5 Flats and 5 F.Darks all in Nikons RAW format (36MPixels each)
Going to preprocess them without any binning leads to a "out of memory" error. I am wondering on a 64Bit system with 48 GByte of memory ?
With 2x2 binning it worked.
According to my understanding on optics and physics I have an optical system that produces PSF with around 10 micons
in diameter that will be sampled by the D800 with 4.8 mycron pixels - a perfect match on sampling according to Nyquist.
So why should I do a 2x2 binning which would lead to a strong undersampling ?
Do I have a knot in my thoughts? And how can make AA5 to work with my resolution ?
Best Regards
Klaus
Beginner's Problem to understand AA5
-
Forum_2014
- Posts: 253
- Joined: 03 Dec 2018, 22:33
Re: Beginner's Problem to understand AA5
Hi, if you are using Windows 7/8 64 bit you can solve installing the prerelease of the Service Pack 4 (please contact me by email) which is currently under development. It supports "Large Address Aware" ( https://www.google.com/search?hl=en&q=l ... ress-aware ).
At 36 megapixels (color) per image it was not possible to have five "streams" at the same time (result, lights, darks, flats, f.darks) but you may solve disabling the f.darks and creating a master flat by hand (then save it) before launching preprocessing. The number of images per stream does not matter.
Seeing usually limits the PSF to 20 micron, so if you work in binning 2x2 you may not lose anything. You can verify that easily: if your stars are bigger than 4-5 pixels then your images are oversampled and you can work in binning 2x2 safely.
Kind regards,
Fabio.
At 36 megapixels (color) per image it was not possible to have five "streams" at the same time (result, lights, darks, flats, f.darks) but you may solve disabling the f.darks and creating a master flat by hand (then save it) before launching preprocessing. The number of images per stream does not matter.
Seeing usually limits the PSF to 20 micron, so if you work in binning 2x2 you may not lose anything. You can verify that easily: if your stars are bigger than 4-5 pixels then your images are oversampled and you can work in binning 2x2 safely.
Kind regards,
Fabio.
-
Forum_2014
- Posts: 253
- Joined: 03 Dec 2018, 22:33
Re: Beginner's Problem to understand AA5
Hi Fabio,
thank you for the explanations. Yes I am using Windows 7 64bit and I am very interested on the prerelease of the SP 4. I have send you a PN with my
email address.
My faint stars have at Half Hight around 3 pixels. The bright one of course more. So from that point of view I am a little bit oversampled.
With 200mm focal lenght I believe that seeing is not of high importance to the width of the PSF. My next tests will show and improve my learning.
I just have restarted my astrophotography adventure after a pause of 20 years and need to change my mindset from the old days of Technical Pan
to the modern sensors.
Have a nice weekend
Klaus
thank you for the explanations. Yes I am using Windows 7 64bit and I am very interested on the prerelease of the SP 4. I have send you a PN with my
email address.
My faint stars have at Half Hight around 3 pixels. The bright one of course more. So from that point of view I am a little bit oversampled.
With 200mm focal lenght I believe that seeing is not of high importance to the width of the PSF. My next tests will show and improve my learning.
I just have restarted my astrophotography adventure after a pause of 20 years and need to change my mindset from the old days of Technical Pan
to the modern sensors.
Have a nice weekend
Klaus
