Bug: area selection changes with Smart Histogram measurement

Discussions, Bug Reports and Issues related to Beta versions of SharpCap
User avatar
Menno555
Posts: 1053
Joined: Mon Apr 20, 2020 2:19 pm
Location: The Netherlands
Contact:

Bug: area selection changes with Smart Histogram measurement

#1

Post by Menno555 »

Version 4.0.7390.0

I select an area to measure with the Smart Histogram. See screenshot 1.
I click Measure, it starts and as soon as the next measurement capture starts, the area selection becomes small and is set to the left side, see screenshot 2.
Restarted SC, the same happens.

Menno
Screenshot 1
Screenshot 1
Screenshot1.jpg (476.18 KiB) Viewed 1329 times
Screenshot 2
Screenshot 2
Screenshot2.jpg (424.37 KiB) Viewed 1329 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: Bug: area selection changes with Smart Histogram measurement

#2

Post by admin »

Hi,

yes, this is new behaviour in SharpCap 4.0 - it now selects the darkest part of the frame automatically for you (if you look at the 'Brain' window you can see that the instruction text has changed to reflect this new behaviour).

cheers,

Robin
User avatar
Menno555
Posts: 1053
Joined: Mon Apr 20, 2020 2:19 pm
Location: The Netherlands
Contact:

Re: Bug: area selection changes with Smart Histogram measurement

#3

Post by Menno555 »

Ah, okay. A typical case of RTFM on my side :oops:
But It doesn't matter that there is a somewhat bright star in the measurement area?

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

Re: Bug: area selection changes with Smart Histogram measurement

#4

Post by admin »

Hi Menno,

shouldn't do - it takes the median brightness of the area, so bright pixels will have no effect unless they are covering a big fraction of the area (if we were using the mean brightness then it would be a problem - hence using median). The selected area is chosen by dividing the frame into 100 sub areas and picking the darkest one.

cheers,

Robin
User avatar
Menno555
Posts: 1053
Joined: Mon Apr 20, 2020 2:19 pm
Location: The Netherlands
Contact:

Re: Bug: area selection changes with Smart Histogram measurement

#5

Post by Menno555 »

Okay, totally clear, Robin. Thanks.
But still a kind of mini text-bug then: the text in SC mentions "... the darkest are in view."

Menno
User avatar
Menno555
Posts: 1053
Joined: Mon Apr 20, 2020 2:19 pm
Location: The Netherlands
Contact:

Re: Bug: area selection changes with Smart Histogram measurement

#6

Post by Menno555 »

Hi Robin

I tried this all last night and a question and a observation.

Question (based upon the observation below): in this case the testing area was closely hugging the edge of my FOV. When it measures, does aberrations, vignetting, and so on, matter?

Observation: last night I was busy with NGC 3690. With the previous version of SC Beta, in that exact same FOV, I choose my own dark area and measured and it gave a solution of 411 seconds, Gain 10 and Offset/Brightness 4. Works great, captures are perfect and still are.
Last night I did try the new measurement. Equipment and so on were exactly the same, including the filter: Baader Skyglow filter, arather transparent broadband filter. Circumstances ofcourse different: moon, skyglow, and so on. By now I now that the Smart Histogram solution will differ because of that but that's most of the time like 30 seconds more or less, Gain 5 more or less, and so.
But now (with the exact same values filled into the Smart Histogram before measuring) it would not give a solution. Only when I got past 10 minutes exposure time, it started to give solutions. This (ofcourse) will not work here in my Bortle 7/8 skies.
Also I never had that before. Solutions in general give max 420 seconds, it never gave me a "There is no solution for this exposure time". Not even with my Optolong L-eNhance filter (3 band narrow)
User avatar
admin
Site Admin
Posts: 13177
Joined: Sat Feb 11, 2017 3:52 pm
Location: Vale of the White Horse, UK
Contact:

Re: Bug: area selection changes with Smart Histogram measurement

#7

Post by admin »

Hi Menno,

thanks, typo corrected.

In theory it's good to measure at the edge of the FOV - if you have vignetting then the illumination there is dimmer meaning a longer exposure is needed. Of course you may want to ignore that in practice ... hmmm.

On the longer exposures, SharpCap previously calculated on the green channel for colour cameras - now it does so on the darkest of the three channels (almost certainly blue). Again probably a good thing technically, but is it the right answer. Thoughts welcome.

cheers,

Robin
User avatar
Menno555
Posts: 1053
Joined: Mon Apr 20, 2020 2:19 pm
Location: The Netherlands
Contact:

Re: Bug: area selection changes with Smart Histogram measurement

#8

Post by Menno555 »

Thanks for the explanation, Robin.

I do know a lot of people will crop their captures, removing a bit or lot from the edge because of aberrations or vignetting. Maybe it's an idea to make an option for a manual measurement area?

My Zwo ASI071MC Pro indeed has blue as the darkest color. I don't know about the technical stuff though :)
I do know that although maybe technically correct, the solutions the Smart Histogram are giving here now are way beyond my scope (pun intended :P).
I tried it again just now in a random piece of sky at zenith and again the solution gives a way longer exposure time than that I can do here with the light pollution. Maybe it works for a dark sky but 10 minutes or more are just not possible with light pollution. Maybe with a narrowband filter but not with a general light pollution filter.

Also, in this measurement, is the white balance of any influence? If so, then the default of the 071 (R60/B99) could help to shorten the solution. But it's now at 50/50 because that gives better results in processing and photometric color calibration.
User avatar
admin
Site Admin
Posts: 13177
Joined: Sat Feb 11, 2017 3:52 pm
Location: Vale of the White Horse, UK
Contact:

Re: Bug: area selection changes with Smart Histogram measurement

#9

Post by admin »

Hi Menno,

white balance controls don't really help - they amplify the signal after it has been detected rather, so will not affect the brain calculations.

Fair enough - I will add some options to the measurement window to allow the green channel to be used and also to restrict to the center region of the frame in some way.

cheers,

Robin
neverfox
Posts: 12
Joined: Fri Feb 05, 2021 6:39 pm

Re: Bug: area selection changes with Smart Histogram measurement

#10

Post by neverfox »

Same issue. I was getting great and reasonable recommendations from the Brain with v3 but v4 just goes and goes and goes. It was trying exposure times up to 2000s before I give up (with the "darkest area" being at the edge or corner where I have vignetting. The Brain was something I used to use every night as a way to check that my exposure time made sense under the conditions but now it's impractical (and likely wrong). Going to revert to 3.3 beta in the meantime (as I'm using an ASI294MM Pro). I'll be following this topic closely.
Post Reply