Brain recommendations when min exposure changes

Somewhere to ask questions about the best way to use SharpCap
Forum rules


If you have a problem or question, please check the FAQ to see if it already has an answer : https://www.sharpcap.co.uk/sharpcap-faqs
Post Reply
zerolatitude
Posts: 51
Joined: Mon Mar 01, 2021 5:24 am

Brain recommendations when min exposure changes

Post by zerolatitude »

Hi,

Following is sequence of steps which should be reproducible by setting the sky rate manually
1. Set Brain capture settings to default - 0.1s min exposure, max 2min, RN limit 5%, channel Green, total stack time 60min
2. Measure sky electron rate with SC. Comes to 16 e/s/px
3. Brain shows Recommended minimum exposure 2.9s, Gain 100, Offset 5

That results in too many subs, so I set minimum exposure to 15s.

Now, the recommended sub is 19.0s (instead of 2.9s), and Gain 0 (instead of 100).

Why would this happen since the read noise, RN limit, and sky rate are the same as earlier?
Attachments
Brain_2.png
Brain_2.png (52.2 KiB) Viewed 305 times
Brain_1.png
Brain_1.png (52.88 KiB) Viewed 305 times
User avatar
admin
Site Admin
Posts: 6621
Joined: Sat Feb 11, 2017 3:52 pm
Location: Vale of the White Horse, UK
Contact:

Re: Brain recommendations when min exposure changes

Post by admin »

Hi,

so I can tell from your graphs that you are using a camera that has a LCG/HCG (low conversion gain/high conversion gain) switchover - probably at gain 60 or close to that point - that shows up in the sensor graphs as a dramatic drop in the read noise as the gain goes past that level.

Once you move the gain past the read noise reduction, much shorter exposures become possible without read noise contributing to the final stacked image. These individually have less dynamic range each that an exposure at minimum gain, but you can take more of them in an hours imaging, so the dynamic range of the final stack is greater using the large number of 2.9s exposures at 100 gain compared to a smaller number of 19s exposures at minimum gain. Since you have set up the brain to aim for maximum dynamic range, it suggests this option.

Now you increase the minimum exposure to 15s - You can no longer take that large number of 100 gain exposures in 1 hour - the number of exposures that can be taken drops from ~1240 to 240. That means that the 100 gain option is no longer the best choice, so SharpCap finds the best results that do fit in with the minimum exposure length, which means recommending minimum gain @19s in your case.

Effectively by setting the minimum exposure length, you have prevented one of the big advantages of short exposure high gain imaging from being exploited (ie you can no longer take a very large number of frames), meaning that the best results are now found elsewhere.

cheers,

Robin
zerolatitude
Posts: 51
Joined: Mon Mar 01, 2021 5:24 am

Re: Brain recommendations when min exposure changes

Post by zerolatitude »

Thanks.

So basically , for a given integration time, since the maximization is for the DR of the stack, and the number of subs is reducing, the DR per sub needs to go up, so higher exposure. Is that right?

There is still a practical / communication issue.

As per the equations in your post on optimum sub exposure, video on choosing the right gain, as also the Brain output, with default settings I got 2.9s as the minimum sub time and 100 as the recommended gain.

I could quite happily have said oh 2.9 is minimum, so 15 will be slightly better, and never thought about changing the gain. This came up because I just happened to put in 15 as the minimum exposure.

I noticed this happens only when aiming for Max DR. When choosing Unity it doesn't change, but then unity doesn't mean much by itself.

So what should be done? Should each planned sub time be tested against the Brain when using Max DR? Is this specific to cameras with dual modes?

If so, would be useful if the Brain showed that when choosing Max DR.

Thanks
User avatar
admin
Site Admin
Posts: 6621
Joined: Sat Feb 11, 2017 3:52 pm
Location: Vale of the White Horse, UK
Contact:

Re: Brain recommendations when min exposure changes

Post by admin »

Hi,

I think the important thing here is that if you go longer than the suggestion then it *may* reduce the maximum dynamic range of the stack. However the max dynamic range is very much a secondary measure of image quality - the primary measure is stack noise level, which is not adversely affected by going for longer exposures.

The difficulty in showing more info on this to the user is that the 'brain' window is already very information dense and geeky - adding more text/information/warning info may just mean that people read less of it. I think that this is probably something to note in the documentation at the next update.

cheers,

Robin
zerolatitude
Posts: 51
Joined: Mon Mar 01, 2021 5:24 am

Re: Brain recommendations when min exposure changes

Post by zerolatitude »

When looking at the data I posted, the differences in DR is maybe 0.5 at most. I doubt it will be perceptible.

I also tried a 30 min stack of 15s@G100 vs 30s@G0. The 30s was much noisier, lijely because my LP is very high, and maybe also because of the smaller number of subs in the stack.

So for actual imaging, will probably stick to G100 and shorter exposures. It was good to understand whats happening though.



Thanks.
User avatar
oopfan
Posts: 996
Joined: Sat Jul 08, 2017 2:37 pm
Location: New York
Contact:

Re: Brain recommendations when min exposure changes

Post by oopfan »

Zero,

The reason why the G0 stack is noisier than the G100 stack is due to higher Read Noise. In my opinion, there are only two reasons to image at G0:
1. You are performing Photometry.
2. You are doing AP, but the target has a high dynamic range like M42 Orion Nebula.

EDIT: Other HDR targets include Globular Clusters where you want to capture light from faint stars without blowing out the core.

Brian
zerolatitude
Posts: 51
Joined: Mon Mar 01, 2021 5:24 am

Re: Brain recommendations when min exposure changes

Post by zerolatitude »

Brian,

Thanks for the reply.

I used G0 (as a trial) only because the Brain recommended it at a 15s min exposure.

The optimal exposure suggested is supposed to be the point at which RN is swamped by skynoise.

At G100, RN gets swamped at 2.9s, so definitely swamped at 15s.

At G0, RN gets swamped at 18s, so definitely at 30s

so really, from the RN point of view, there shouldn't be a difference between the two.

Is that wrong?
Post Reply