Author Topic: Histogram question/problem  (Read 2449 times)

Offline tdgm

  • PixInsight Old Hand
  • ****
  • Posts: 302
Histogram question/problem
« on: 2018 February 15 10:27:37 »
Hello Everyone

Recently I sent a .jpg to a friend who also uses PI. When he opened the program the Histogram showed him some clipping of the black point by the left edge of the Histogram but on my computer it does not show clipping. Below is a screenshot of both machines on the top is my friends computer with t he image and a Histogram and on the bottom is my machine. The question is why are the Histograms different for the same image. He changed the name for his file system. I apologize if image is a little small.

Offline pfile

  • PTeam Member
  • PixInsight Jedi Grand Master
  • ********
  • Posts: 4729
Re: Histogram question/problem
« Reply #1 on: 2018 February 15 11:07:36 »
ICC profiles perhaps? what happens if you strip any ICC profiles from the images before loading them?

rob

Offline tdgm

  • PixInsight Old Hand
  • ****
  • Posts: 302
Re: Histogram question/problem
« Reply #2 on: 2018 February 15 11:44:06 »
How would I check that?

Steve

Offline oldwexi

  • PixInsight Guru
  • ****
  • Posts: 627
    • Astronomy Pages G.W.
Re: Histogram question/problem
« Reply #3 on: 2018 February 15 12:02:50 »
You have checked
"Reject saturated Pixels for histogram representation"
and your friend did not.
This can give different histogram views as you both did not
select the same histogram representation.

Gerald

Offline msmythers

  • PTeam Member
  • PixInsight Jedi
  • *****
  • Posts: 1178
    • astrobin
Re: Histogram question/problem
« Reply #4 on: 2018 February 15 12:18:11 »
Looking at the screenshoot I think Gerald is right on the mark. It looks like the rejected pixels is not selected in the upper histogram tool

Here is an example with the rejected pixels selected and then deselected.


Mike

Offline tdgm

  • PixInsight Old Hand
  • ****
  • Posts: 302
Re: Histogram question/problem
« Reply #5 on: 2018 February 15 12:29:58 »
OK I see what you guys are suggesting but when I check and uncheck rejected saturated pixels my image Histogram(bottom) does not change. What I was looking at when I put this situation up was that in the bottom image and graph there is a toe and the other one does not have a toe. Or another way of saying this is the upper histogram shows clipping of the black point and the bottom one does not.

Steve

Offline msmythers

  • PTeam Member
  • PixInsight Jedi
  • *****
  • Posts: 1178
    • astrobin
Re: Histogram question/problem
« Reply #6 on: 2018 February 15 12:51:45 »
Well now that rejected pixels have been eliminated did you email the image Steve or was this on a hosting site? Some hosting sites do subtle manipulations. Just a shot in the dark. Can your friend email back the image for you to check?


Mike   

Offline tdgm

  • PixInsight Old Hand
  • ****
  • Posts: 302
Re: Histogram question/problem
« Reply #7 on: 2018 February 15 12:55:23 »
Hi Mike


Image was emailed directly


Steve

Offline tdgm

  • PixInsight Old Hand
  • ****
  • Posts: 302
Re: Histogram question/problem
« Reply #8 on: 2018 February 15 19:38:39 »
Mike Gerald and Rob

I have figured out the problem it was in the email program I sent the image with. Since the file size was 4.8 mb my Mac maill program I decreased the size to 299 kb and the size of image went down from 2750/2200 to 1200/960. When I received the image back  and checked the original sent image I noticed the size difference. When checking the histogram on  the smaller image I saw the histogram looked k=like some clipping was going on.


Thanks for thew suggestions


Steve

Offline pfile

  • PTeam Member
  • PixInsight Jedi Grand Master
  • ********
  • Posts: 4729
Re: Histogram question/problem
« Reply #9 on: 2018 February 15 22:10:31 »
ah - yeah Mail.app has those image rescaling settings. good catch.

rob

Offline msmythers

  • PTeam Member
  • PixInsight Jedi
  • *****
  • Posts: 1178
    • astrobin
Re: Histogram question/problem
« Reply #10 on: 2018 February 16 00:29:54 »
Steve

Glad the mystery is solved.  :)


Mike