sticky exposure length or sticky gain/offset for sky-limited exposures?

Discussion of using SharpCap for Deep Sky Imaging
mAnKiNd
Posts: 303
Joined: Sun Jun 11, 2017 4:55 pm
Location: Long Island, NY
Contact:

Re: sticky exposure length or sticky gain/offset for sky-limited exposures?

Post by mAnKiNd » Wed Sep 12, 2018 9:58 pm

It is also important I feel to ask what do we define as our target in our optimised livestack? Is it a target SNR value? Bit-depth? Or is it not empirical and simply an aesthetic judgement?

User avatar
oopfan
Posts: 684
Joined: Sat Jul 08, 2017 2:37 pm
Location: New York
Contact:

Re: sticky exposure length or sticky gain/offset for sky-limited exposures?

Post by oopfan » Wed Sep 12, 2018 10:22 pm

I always liked the idea of keeping gain, offset, and exposure the same. Not only do I get to reuse darks but I also get to set the exposure based on maximum light using the luminance filter. Having a full-well depth of a mere 15000 electrons makes me painfully aware of accidentally saturating stars. So when I set the initial exposure with the luminance filter I am confident that RGB will not cause problems.

Brian

mAnKiNd
Posts: 303
Joined: Sun Jun 11, 2017 4:55 pm
Location: Long Island, NY
Contact:

Re: sticky exposure length or sticky gain/offset for sky-limited exposures?

Post by mAnKiNd » Wed Sep 12, 2018 10:45 pm

oopfan wrote:
Wed Sep 12, 2018 10:22 pm
I always liked the idea of keeping gain, offset, and exposure the same. Not only do I get to reuse darks but I also get to set the exposure based on maximum light using the luminance filter. Having a full-well depth of a mere 15000 electrons makes me painfully aware of accidentally saturating stars. So when I set the initial exposure with the luminance filter I am confident that RGB will not cause problems.

Brian
This is a wise and elegant approach.

mAnKiNd
Posts: 303
Joined: Sun Jun 11, 2017 4:55 pm
Location: Long Island, NY
Contact:

Re: sticky exposure length or sticky gain/offset for sky-limited exposures?

Post by mAnKiNd » Thu Sep 13, 2018 12:45 am

..though I'm not sure how one could keep the narrowband channel at the same exposure & gain/offset as the RGB channels and still get a decent single exposure histogram. Perhaps, treat the narrowband channel differently. Max DR for LRGB channels and a higher gain for the narrowband channels.

If it's kept at the same exposure settings as LRGB, even with a 'weaker' histogram, would it be simply the case of obtaining a much larger number of frames in the stack to make up for it?

Cheers

mAnKiNd
Posts: 303
Joined: Sun Jun 11, 2017 4:55 pm
Location: Long Island, NY
Contact:

Re: sticky exposure length or sticky gain/offset for sky-limited exposures?

Post by mAnKiNd » Thu Sep 13, 2018 12:52 am

Or perhaps simply just use the smart brain on each channel for finding the max DR sky limited exposures, acquire calibration files and simply stop looking for unified settings, base it all on total integration time.

User avatar
oopfan
Posts: 684
Joined: Sat Jul 08, 2017 2:37 pm
Location: New York
Contact:

Re: sticky exposure length or sticky gain/offset for sky-limited exposures?

Post by oopfan » Thu Sep 13, 2018 1:17 am

Yes, The Brain would work but I would be sure to dial in some Offset and take a test snapshot to make sure that stars aren't saturating.

The tricky part is Live Stack and how often you should tell it to auto-save. For me I know what the proper color balance is for my RGB filters but I couldn't tell you what it should be for SHO. Maybe to begin with on the first night save every frame like I do without using Live Stack. Then go indoors and process it in order to find the right mix. With that knowledge you can use Live Stack on subsequent nights.

I've decided that my next challenge will be SHO of a part of the Soul Nebula, specifically here.

My sensor's FOV closely approximates hers. It's nearly exact in the vertical dimension but my width is wider.

Brian

mAnKiNd
Posts: 303
Joined: Sun Jun 11, 2017 4:55 pm
Location: Long Island, NY
Contact:

Re: sticky exposure length or sticky gain/offset for sky-limited exposures?

Post by mAnKiNd » Thu Sep 13, 2018 1:41 am

Brian thanks.

It's my understanding that if I were to use the smart brain (whose exposure setting analysis is based by default for a 1 hour stack) to figure out the best exposure & gain for each channel, say to achieve max. dr, then with that 'equilibrium between channels' in place, one could simply rely on total integration time for the stack, say auto-save every 60 minutes (which goes nicely with the smart brain analysis). Then, using 1 hour stacks from each night, double stack them for each channel.

Thanks again
Minos

mAnKiNd
Posts: 303
Joined: Sun Jun 11, 2017 4:55 pm
Location: Long Island, NY
Contact:

Re: sticky exposure length or sticky gain/offset for sky-limited exposures?

Post by mAnKiNd » Thu Sep 13, 2018 1:42 am

oopfan wrote:
Thu Sep 13, 2018 1:17 am

I've decided that my next challenge will be SHO of a part of the Soul Nebula, specifically here.

My sensor's FOV closely approximates hers. It's nearly exact in the vertical dimension but my width is wider.

Brian
That sounds awesome! Go for it :)

User avatar
oopfan
Posts: 684
Joined: Sat Jul 08, 2017 2:37 pm
Location: New York
Contact:

Re: sticky exposure length or sticky gain/offset for sky-limited exposures?

Post by oopfan » Thu Sep 13, 2018 1:48 am

Several months ago during winter I imaged the Crab Nebula in Ha using The Brain here. Unfortunately I didn't dial in Offset. The stack was unbelievably noisy with lots of raining noise. This, I am certain, was caused by incomplete dark frame subtraction due to clipping on the LHS of the histogram. Anyhow I was finally able to process it but it was a struggle. I never want to go through that again. Lesson learned: always dial in some Offset.

Today I looked at the light frames in FITS Liberator. The brightest stars (9th magnitude) came in at 25000 out of 65535, so there was no saturation. This was at Unity Gain (i.e. 1.0 e-/ADU). The Brain recommended an exposure of 46 seconds. I followed it.

Here is my plan for the Soul Nebula. Who knows when the next clear night will be so I will immediately begin collecting darks at these settings: Gain 389 (i.e. unity gain), Offset 30, and Exposure 50s. I am confident that this will yield usable results. I am going to collect as much data as I can before it crosses the meridian. I figure about 2 hours the first night with Ha. The second night I'll do another 2 hours in SII, and the third night in OIII. It will be interesting to see. There is no way that I am coming close to the quality of Sara Wager's image but I might be pleasantly surprised.

Brian

mAnKiNd
Posts: 303
Joined: Sun Jun 11, 2017 4:55 pm
Location: Long Island, NY
Contact:

Re: sticky exposure length or sticky gain/offset for sky-limited exposures?

Post by mAnKiNd » Thu Sep 13, 2018 12:55 pm

Thanks Brian,

Dialing in offset is indeed important, I ran into the same issue once where I didn't have any offset and my calibration frames did not work properly either. I noticed on my last run of the smart brain analysis that it automatically pushed my offset value to max, perhaps to avoid this shortcoming.

Even at unity gain, your exposure seems it had plenty of headroom (almost 2/3rds).

I'm leaning towards the following now for my project:

- Use smart brain to determine sky brightness levels with my L filter in a dark part of the sky next to the Heart nebula.

- Slew to the Heart & plate solve and then resume smart brain analysis for my 5 filters using either unity gain or max dynamic range (still trying to decide) for a 60 minute live stack.

- Round up the suggested exposure times i.e. 49.3s -> 50s, set the gains to recommended values

- and then proceed to take calibration frames.

- Once all filter channels are setup and calibrated, they will be saved individually as presets for quick selection, along with two extra high gain 2s exposure presets for focusing my LRGB and HO filters.

- Finally, begin acquiring live stacks set to auto-save every 60 minutes.

I think this will work well, but I need to decide whether to set the smart brain to analyse for Unity gain or Max DR. Also need to consider max read noise contribution, by default set at 10%. I suppose if you lower that limit, it will use higher gains.

Your plan for the Soul sounds solid, perhaps when we are both done we can put them together side by side as two approaches to imaging the Heart n' Soul :)

Cheers
Minos

Post Reply