Jump to content

  •  

CNers have asked about a donation box for Cloudy Nights over the years, so here you go. Donation is not required by any means, so please enjoy your stay.

Photo

Monochrome newbie questions: LRGB process, capture settings

Astrophotography Beginner Filters Imaging
  • Please log in to reply
3 replies to this topic

#1 barnold84

barnold84

    Mariner 2

  • -----
  • topic starter
  • Posts: 209
  • Joined: 15 Aug 2020
  • Loc: Germany

Posted 03 March 2021 - 12:01 PM

Hi,

 

Yesterday I tried to experiment with my new mono camera and the LRGB filters to gather some experience with the new equipment. My current experience allowed me to capture some nice (in my opinion) DSO pictures with my stock DSLR and William Optics refractor. On the other hand, I have the new camera and also an Meade 8" ACF that I would like to work with. With my post, I'd like to get some feedback on where I'm likely on the right track and where I am wrong.

 

The equipment:

Scope: Meade 8" ACF f/10 with reducer, yielding an f/6,3.

Camera: fan cooled Altair Hypercam 183M (mono) with pixel size 2,4um.

Guiding: Guide scope. Guiding itself went perfect, star quality on the subs is very good.

 

My target was M81 (Bode's galaxy) as I had just made images of M37 under the bright moon (follow the AstroBin link in my signature). My goal was to capture before the moon rises high above the horizon and use the dark sky.

 

I took 84 L subs with 30 seconds exposure. The RGBs have each 64 subs also at 30 seconds. For all filters, I used 2x2 binning.

 

Here's a list of questions:

1. I've read people take L subs with low binning, even down to 1x1 and RGB at 2x2 or 3x3. With the optics above, a 1x1 binning would have given me 0.4"/pixel. Hence, I didn't see a need to capture L and RGB at different binning. What's your opinion or suggestion here?

 

2. I'm not sure about the exposure times. With the DSLR I took the approach to try to expose as much as possible until the first stars start to saturate, given that the histogram "peak" is quite a bit above the dark noise. I tried doing the same with the mono (12-bit btw. vs. 14-bit on DSLR). The largest stars just saturated a bit and the histogram mean was just above the darks mean. In post processing, it looks to me that I have to stretch too much to start to see details and then the noise comes out. I guess I'd need to expose longer or capture more subs?

 

3. The moon: I started at around 7:30pm. At my place it's just about to get "astronomically" dark. However, I can see on my L subs that mean intensity decreases over time. The R and G have a constant mean. When the moon was about to get above horizon, I can see how the mean intensity of the B subs starts to increase. When I try to compose RGB, the resulting image has a rather strong blue tone. I would attribute this to the moon?

 

4. To compose RGB, I used Affinity Photo. To adjust for the different sensor QE, I am adjusting the brightness of the R and B channel. The result seems to give a rather good color. Your opinion on this procedure?

 

You can see, I have quite some questions. If somebody likes, I can provide the stacked images for each channel so that you could take a look and maybe give a recommendation on which knob I should turn most.

 

Clear Skies!


Edited by barnold84, 03 March 2021 - 12:01 PM.


#2 fewayne

fewayne

    Apollo

  • *****
  • Posts: 1,253
  • Joined: 10 Sep 2017
  • Loc: Madison, WI, USA

Posted 03 March 2021 - 01:22 PM

So long as you're not black-clipping on the left of the  histogram, your subexposures will be sufficient. Perhaps not perfectly optimal WRT read noise, but what really counts is total integration time. What sort of gain were you using?  ZWO did some experiments and recommend that, in general, you should use low gain settings when you need to optimize for dynamic range, and otherwise use high gain to minimize read noise. Unity gain yielded poorer, neither-fish-nor-fowl results.

 

Blue: I made the same  mistake last  night, running my LRGB frames in that order, with moonrise about halfway through the G. A full moon lights the sky blue for the same reason  that the sun does (it's just sunlight on the same  atmosphere, after all). If the moon will rise  late in your sequence, do B first. My impulse would be to shoot L last, reasoning that it's least affected, but that's pretty shaky (would love to be further educated there!).

 

Binning: There's little advantage to binning with a CMOS camera -- unlike CCDs, it's not done at the sensor. You might as well run at full resolution and downsample if needed in  processing, unless it's the file size that appeals.



#3 barnold84

barnold84

    Mariner 2

  • -----
  • topic starter
  • Posts: 209
  • Joined: 15 Aug 2020
  • Loc: Germany

Posted 03 March 2021 - 01:42 PM

Thank you for your reply! I'm adding my responses in the quote:

 

So long as you're not black-clipping on the left of the  histogram, your subexposures will be sufficient. Perhaps not perfectly optimal WRT read noise, but what really counts is total integration time. What sort of gain were you using?  ZWO did some experiments and recommend that, in general, you should use low gain settings when you need to optimize for dynamic range, and otherwise use high gain to minimize read noise. Unity gain yielded poorer, neither-fish-nor-fowl results.

 

I don't want to use unity gain in the wrong meaning, so regarding gain: the setting people are recommending is gain=400, where 1 ADU is 1 electron.

Edit: so, my personal expectation is to go for many more subs. How about the exposure times for all channels? I forgot to say that I am not trying to "over optimize", i.e. if the stars would touch saturation at 33s exposure time, I would set it to 30. 

 

Blue: I made the same  mistake last  night, running my LRGB frames in that order, with moonrise about halfway through the G. A full moon lights the sky blue for the same reason  that the sun does (it's just sunlight on the same  atmosphere, after all). If the moon will rise  late in your sequence, do B first. My impulse would be to shoot L last, reasoning that it's least affected, but that's pretty shaky (would love to be further educated there!).

 

You're right about the scattering of blue light (I should know better...). Continuing from the physics here: blue is scattered the most. The scattering intensity goes by the fourth power of the frequency or inverse wavelength (as you said: the reason why the sky is blue). Hence, I think the best would be: L, B, G, B if one wan't to make the race against a rising moon. Reverse sequence if moon is setting. L covers the blue spectrum and the rest. My reasoning: our perception is more sensitive to intensity than color contrast. Therefore, L quality should be superior tho color quality. So L first (rising moon). Then B as blue will capture all the scattered light from the atmosphere. Green is scattered less (by a factor of 2) and R should be the least responsive.

 

I think you've a good point here.

 

The following plot shows the median intensity (ADUs) for each sub. The frames indicate the filter while I chose yellow for the L. The time is UTC. I'm in UTC+1. You can see the darkening sky in the Ls and the rising moon in the Bs.

 

Untitled.jpeg

 

Binning: There's little advantage to binning with a CMOS camera -- unlike CCDs, it's not done at the sensor. You might as well run at full resolution and downsample if needed in  processing, unless it's the file size that appeals.

 

According to Altair, the binning is an averaging of the pixels of the binned ones. So this is a way to reduce signal to noise ratio. I shot with my DSLR some time ago on the SCT (pixel size 3.8um) and the stars are cotton balls. Something that doesn't look nice and requires downsampling later one. I'd guess it would be fine to save the disk space and just go with the binned subs!?


Edited by barnold84, 03 March 2021 - 01:55 PM.


#4 barnold84

barnold84

    Mariner 2

  • -----
  • topic starter
  • Posts: 209
  • Joined: 15 Aug 2020
  • Loc: Germany

Posted 04 March 2021 - 09:58 AM

Ok, I double checked my RGB exposures. The mean values aren't sufficiently above the darks mean values. Way too much noise. The lights look better but probably I should get more lights and also do the RGB again.




CNers have asked about a donation box for Cloudy Nights over the years, so here you go. Donation is not required by any means, so please enjoy your stay.


Recent Topics





Also tagged with one or more of these keywords: Astrophotography, Beginner, Filters, Imaging



Cloudy Nights LLC
Cloudy Nights Sponsor: Astronomics