Failed Brain Exposure Calc: No Solution

Anything that doesn't fit into any of the other forums
Post Reply
dts350z
Posts: 39
Joined: Fri May 18, 2018 6:00 am

Failed Brain Exposure Calc: No Solution

#1

Post by dts350z »

Ha SharpCap.png
Ha SharpCap.png (841.82 KiB) Viewed 2709 times
Spent > 45 minutes getting the Brain to recommend an Ha exposure. Then was left with "No Solution".

Successfully got it work on a Galaxy with the Lum filter, but when I switched to a Nebula and my Ha filter it took multiple 8 min exposures measuring the sky background and then gave me the no solution.

Any ideas?

I'll go find the log file.
dts350z
Posts: 39
Joined: Fri May 18, 2018 6:00 am

Re: Failed Brain Exposure Calc: No Solution

#2

Post by dts350z »

Log_2018-06-03T21_53_42-13028.log
(481.95 KiB) Downloaded 171 times
User avatar
admin
Site Admin
Posts: 13177
Joined: Sat Feb 11, 2017 3:52 pm
Location: Vale of the White Horse, UK
Contact:

Re: Failed Brain Exposure Calc: No Solution

#3

Post by admin »

Hi,

what that is telling you (in a not very helpful way) is that the 10 minute sub exposure length is not long enough to give optimal results - this is due to the very low sky brightness due to the Ha filter.

What I need to do is

a) allow 15min, 20min, 30min as maximum sub length options
and
b) give better messaging when the sub exposure range chosen stops there from being a good solution.

For now, what you know is that for Ha the optimum sub length is >10min, so it is probably limited by your tracking/guiding.

cheers,

Robin
dts350z
Posts: 39
Joined: Fri May 18, 2018 6:00 am

Re: Failed Brain Exposure Calc: No Solution

#4

Post by dts350z »

OK thanks.

Please also increase the maximum integration time in the Brain to like 12 hours or more.

I'm doing multiple nights on the same target and need 10+ hours to pull detail out of the light pollution (at least when doing LRGB/OSC)
dts350z
Posts: 39
Joined: Fri May 18, 2018 6:00 am

Re: Failed Brain Exposure Calc: No Solution

#5

Post by dts350z »

Also,

What I really hoped to get out of this capability in SharpCap is to get an estimate of best exposure without taking multiple long exposures.

With DSLR, I teach people to use the highest available iso, say 12800, and take short (10 seconds to under a couple of minutes) exposures to determine a good histogram (gap on the left, peak in first third), then multiply the exposure time to arrive at the proper exposure at your target iso.

e.g. 60 seconds @ iso12800 is equivalent to 480secs (3 iso stops, 2*2*2=8x) @iso 1600.

That way you don't waste dark time...

To date I have been unable to figure out how to do similar with these cooled cmos astro cams, maybe because of offset (?) or not (until really today) understanding what all is going on inside the camera.

Will SharpCap be able to do that or will the measurement of sky take 2x or more of the resulting exposure recommendation?

Thanks
User avatar
admin
Site Admin
Posts: 13177
Joined: Sat Feb 11, 2017 3:52 pm
Location: Vale of the White Horse, UK
Contact:

Re: Failed Brain Exposure Calc: No Solution

#6

Post by admin »

Hi,

the total integration time is actually a bit of a placebo - it *only* affects vertical scale of the 'Detection Threshold Simulation' graph - ie how faint a source in electrons/pixel/second can be detected above the noise floor. It doesn't change the shape of that graph or the suggested gain/exposure values.

The 'Brain' procedure includes calculations to check that the final exposure/gain/offset combo will have a histogram that is properly separated from the LHS - it includes both the effects of offset and sky/thermal background in this. The choice of exposure is so as to ensure that the shot noise in each sub-exposure overwhelms the read noise. The gain choice is more subtle - you can either pick unity gain (which is simple) or the gain which gives the maximum dynamic range in the final stack. Often (but not always) this is the minimum gain value. For cameras with switchable low/high conversion gain it often turns out to be the gain where the camera switches from low to high conversion gain as the reduction in read noise by making the switch more than compensates for the reduced full well depth.

cheers,

Robin
dts350z
Posts: 39
Joined: Fri May 18, 2018 6:00 am

Re: Failed Brain Exposure Calc: No Solution

#7

Post by dts350z »

OK, not sure I understand the low to high conversion gain or if it applies to zwo asi1600 but:

Is "max DR in the final stack" assuming that there is interesting stuff in all parts of the histogram?

For instance, in NB we probably don't care about stars being clipped (other than bloat) because the color is false (and I've taken to replacing NB stars with RGB stars in post processing anyway), so it seems to me that we should maximize the DR of the small area the peak of the histogram (say the auto stretched area) and not worry about other sections.

Increasing the gain would give you a wider ADU range for the same range of e-?

The practical upper limit on gain would be where you need all 12 bits to represent that range of e- in ADU?

So maybe (for my current Ha Tulip Nebula) around 139 to 200.

Let's say the auto stretch looks awesome and the range of displayed ADU is 111 to 1940 (real world example from last night). That's a range of 1829 ADU. If we convert that back into e- (@ gain 75, 2.09 e-/ADU) we get an e- range of 3826. Now if we look at the effect of gain on that e- range:

Gain, ADU range, bits for range
0 746 10
60 1509 11
61 1526 11
75 1829 11
100 2378 11
139 3495 12
200 7553 13
300 23656 15
400 72784 16
500 229002 18

It looks like a gain of 139 (hmm, unity in this case) would give use the maximum ADU DR for the area of interest?
User avatar
admin
Site Admin
Posts: 13177
Joined: Sat Feb 11, 2017 3:52 pm
Location: Vale of the White Horse, UK
Contact:

Re: Failed Brain Exposure Calc: No Solution

#8

Post by admin »

Max Dynamic Range in Stack is trying to optimize:

ratio of (brightest thing that is not quite saturated) to (final stack noise level in the darkest parts of the image).

If you want to use a different gain to the one suggested for one reason or another you can read the optimum exposure length for any gain off of the graph in the 'Optimal Exposures' tab - just click on the graph at the gain you want to use and it will tell you the exposure time you should aim for.

Not really sure I follow your calculation example I'm afraid.

Robin
Post Reply