Hi Robin,
I got a few questions aiming to optimize my EAA imaging.
First, a few infos regarding my equipment: I‘m using a 10“ f/5 GSO dob reduced to f/3.75 with Starizona Nexus 0.75x CC/FR over a single-axis EQ platform. My skies can be classified as Bortle 6. The camera in use is an uncooled ZWO ASI533 MC, which exhibits very low read noise.
Regarding the darks vs hot/cold pixel removal:
So far, I‘ve used darks and usually take around 25 at my typical settings of 8s exposures at gain 300.
I was wondering whether my low read noise camera would do equally well if I only used the hot/cold pixel removal method? What would you say?
This would be favorable since at the moment, temperature usually drops over the night and the darks should not be as effective then?
Regarding the background subtraction:
Until now I‘ve mainly used „blended offset“, sometimes experience nasty gradients. Given my Bortle 6 skies and a few streetlights behind my house (not in direct sight of the telescope), what mode would you recommend?
Thanks in advance & CS
Johannes
Background subtraction & darks vs. pixel removal
Re: Background subtraction & darks vs. pixel removal
Forgot to mention that I’m using the Baader Neodymium M&SG filter for galaxies, open/globular clusters etc. and the Optolong L-eNhance for emission/planetary nebulae.
How would this affect the choice of background subtraction modes? Would I use the same mode when applying different filters?
How would this affect the choice of background subtraction modes? Would I use the same mode when applying different filters?
-
admin
- Site Admin
- Posts: 13350
- Joined: Sat Feb 11, 2017 3:52 pm
- Location: Vale of the White Horse, UK
- Contact:
Re: Background subtraction & darks vs. pixel removal
Hi,
so you are definitely thinking about the right options - the IMX533 and IMX571 (and related) sensors have very low read noise and also practically no fixed 'amp glow' noise of the sort that caused problems with cameras based on the IMX183. Thermal noise is also low on CMOS cameras compared to older amateur CCD models. That combination makes 'no dark' imaging with just hot pixel removal viable, particularly if you are working with relatively short exposures (up to 30s or so). I would suggest 'Hot and Cold Pixel Removal' is probably the best option, as cold pixels can be just as problematic as hot ones once you start stretching the image a long way.
For the background subtraction, the best bet is usually to use the simplest option that works, so if you have gradients then choose 'Linear Gradient Removal' rather than the non-linear ones as a first try. With the more complex options such as non-linear and low frequency wave, there is a bigger chance of the algorithm incorrectly picking up the target as the background (particularly on a large target that nearly fills the frame).
When you add the filter, evaluate the image without background subtraction to see if the filter has removed enough light pollution to eliminate the gradient (this will depend on the nature of the light pollution and how much of it falls within the cut-out bands of the filter). If it does a good job of reducing the background and the gradient then maybe try with the 'blended offset' - if you can still see the gradient in the background, stick with linear gradient removal.
Hope this helps,
Robin
so you are definitely thinking about the right options - the IMX533 and IMX571 (and related) sensors have very low read noise and also practically no fixed 'amp glow' noise of the sort that caused problems with cameras based on the IMX183. Thermal noise is also low on CMOS cameras compared to older amateur CCD models. That combination makes 'no dark' imaging with just hot pixel removal viable, particularly if you are working with relatively short exposures (up to 30s or so). I would suggest 'Hot and Cold Pixel Removal' is probably the best option, as cold pixels can be just as problematic as hot ones once you start stretching the image a long way.
For the background subtraction, the best bet is usually to use the simplest option that works, so if you have gradients then choose 'Linear Gradient Removal' rather than the non-linear ones as a first try. With the more complex options such as non-linear and low frequency wave, there is a bigger chance of the algorithm incorrectly picking up the target as the background (particularly on a large target that nearly fills the frame).
When you add the filter, evaluate the image without background subtraction to see if the filter has removed enough light pollution to eliminate the gradient (this will depend on the nature of the light pollution and how much of it falls within the cut-out bands of the filter). If it does a good job of reducing the background and the gradient then maybe try with the 'blended offset' - if you can still see the gradient in the background, stick with linear gradient removal.
Hope this helps,
Robin