Histogram question

Somewhere to ask questions about the best way to use SharpCap
cuivenion
Posts: 178
Joined: Wed Feb 15, 2017 7:37 pm

Histogram question

Post by cuivenion » Wed Jan 23, 2019 3:06 pm

I was running some dark frames with ASI224 and experimenting with the brightness level. The brightness level is being used to make sure the histogram doesn't tough the left hand side but when looking at the histogram the horizontal bars show it not touching but the normal histogram looks like it is. Which should I trust?
Without Blue.jpg
Without Blue.jpg (228.62 KiB) Viewed 260 times
Also, on ocassional frames the blue shoots up to 100% on the horizontal bar and then returns to normal again. Is this something to worry about?

cuivenion
Posts: 178
Joined: Wed Feb 15, 2017 7:37 pm

Re: Histogram question

Post by cuivenion » Wed Jan 23, 2019 3:28 pm

Also, over a exposure of 7 seconds (gain 300) the horizontal green bar shoots up to 100% and stays there.

BlackWikkett
Posts: 146
Joined: Sat Jun 23, 2018 8:48 pm
Contact:

Re: Histogram question

Post by BlackWikkett » Wed Jan 23, 2019 3:40 pm

Hi Cuivenion,

This looks quite normal to me. Just make sure your dark frames match the exact settings of your light frames. The signal that the histogram is reflecting is the dark current of the camera. There's amp glow, hot pixels and I wouldn't be surprised if you're seeing a random high energy particle impact on the sensor being reflected. All nothing to worry about. Proof will be when you apply the dark frame to lights.

-Wikkett

User avatar
admin
Site Admin
Posts: 2516
Joined: Sat Feb 11, 2017 3:52 pm
Location: Vale of the White Horse, UK
Contact:

Re: Histogram question

Post by admin » Wed Jan 23, 2019 4:30 pm

Hi,

It's basically hot pixels in the green that is pushing the summary bar out to the right there when you get to a longer exposure. That bar extends to the maximum value for that channel excluding a small number of pixels (about one and a Hundred thousand of the total pixels). As you can see from the green spike on the right-hand side of the histogram you have about 30 or so hot pixels in the green which is enough to trigger the summary bar to move all the way to the right. In this case trust the histogram graphs – the summary bar is just bringing your attention to the fact that there are starting to be a significant number of green pixels that are overexposed.

Cheers, Robin

cuivenion
Posts: 178
Joined: Wed Feb 15, 2017 7:37 pm

Re: Histogram question

Post by cuivenion » Wed Jan 23, 2019 4:57 pm

Ok thanks guys. On the issue of black clipping are the horizontal bars accurate? Is there no black clipping when the horizontal bars are not touching the left hand side? I only ask as the graph can be quite hard to read in that area.

User avatar
turfpit
Posts: 647
Joined: Mon Feb 13, 2017 8:13 pm
Location: UK
Contact:

Re: Histogram question

Post by turfpit » Wed Jan 23, 2019 5:27 pm

Before starting an imaging run, I always take a test frame and scrutinise the Image Statistics in FITS Liberator (don't want to waste an hour collecting duds). In the example below, using an Altair 183C, I wasn't sure about the RHS with the histogram for the actual capture. FITS Liberator confirmed a Max of 65504, which left me confident that the image was not over-exposed.

I always work with the Histogram undocked, as I find the compact display gives a better graphical representation - YMMV.

Dave

FITS-Liberator.JPG
FITS-Liberator.JPG (60.95 KiB) Viewed 246 times

capture-in-progress.JPG
capture-in-progress.JPG (65.24 KiB) Viewed 246 times

cuivenion
Posts: 178
Joined: Wed Feb 15, 2017 7:37 pm

Re: Histogram question

Post by cuivenion » Wed Jan 23, 2019 5:57 pm

Thanks, it's more the black clipping I'm concerned about though. I presume the the Min would be 0 in Fits Liberator if the frame was underexposed, but if the horizontal summary bars in the Sharpcap histogram can be used that would be a lot quicker.

User avatar
oopfan
Posts: 456
Joined: Sat Jul 08, 2017 2:37 pm
Location: New York
Contact:

Re: Histogram question

Post by oopfan » Wed Jan 23, 2019 6:48 pm

One might expect that 65504 to be just below saturation but it appears not to be the case. I have never seen a value over it.

Technically the maximum value ought to be 4095 * 16 = 65520 but I've never seen it.

If saturation is something you want to avoid I recommend keeping the max value below 65000.

Brian

User avatar
turfpit
Posts: 647
Joined: Mon Feb 13, 2017 8:13 pm
Location: UK
Contact:

Re: Histogram question

Post by turfpit » Wed Jan 23, 2019 7:17 pm

I couldn't find a black clipping example in my data, haven't had that for months now. Hence I used the overexposed to point the way to the method I use.
I presume the the Min would be 0 in Fits Liberator if the frame was underexposed
Underexposed or offset not optimal. A way to determine optimal offset was suggested in your recent offset question thread
viewtopic.php?f=9&t=1209
if the horizontal summary bars in the Sharpcap histogram can be used that would be a lot quicker
I wasted too many good nights with 'quick', now I lean towards getting it right. A few minutes to avoid a wasted session is time I consider well spent.

At the end of the day, the definitive answer is to try settings with your equipment & evaluate your results against those settings. Once you know what work then progression follows. It is also important to know what doesn't work.

Dave

cuivenion
Posts: 178
Joined: Wed Feb 15, 2017 7:37 pm

Re: Histogram question

Post by cuivenion » Wed Jan 23, 2019 8:47 pm

Hi there may be some confusion. I was just wondering if the horizontal summary bars accurately show if you're black clipping or not because the graph is quite hard to read that close to the left hand side. The horizontal bars are quite easy to read for that purpose.

Post Reply

Who is online

Users browsing this forum: No registered users and 1 guest