Thresholding bug

The Curvemeister 3 beta test. Please post and discuss bugs, screen shots, suggestions, and any other information about your testing. This board is only visible to Curvemeister 3 beta testers.
ggroess
Posts: 5342
Joined: Wed May 24, 2006 2:15 am
Contact:

Postby ggroess » Sun Nov 09, 2008 5:47 pm

Mike,
The Shadow thresholding does not accurately reflect the grid controls on the image.  See screen shots attached. 
Shot 1 is where the black specks show up on the image.  Showing the histogram and the placement of the threshold.  Image 2 is the result if you place the threshold where the specks show up....

Greg

-default
Posts: 1916
Joined: Thu Mar 26, 2015 1:53 am

Postby -default » Mon Nov 10, 2008 2:28 am

Greg, I think it's probably working correctly - or I'm misunderstanding the problem, which is also likely.

I opened your second image in CM, and clicked on the edge of the Lab curve, and the preview looks about the same - black dots.  Screenshot included.

-default
Posts: 1916
Joined: Thu Mar 26, 2015 1:53 am

Postby -default » Mon Nov 10, 2008 2:34 am

Rereading your post, I'm thinking I totally misunderstood what you were getting at.  Is it that the histogram doesn't seem to match up with the darkest parts of the image?

ggroess
Posts: 5342
Joined: Wed May 24, 2006 2:15 am
Contact:

Postby ggroess » Mon Nov 10, 2008 4:32 am

No, it is that the black threshold is over doing the correction. 
I have to back out to the edge of the histogram to get an acceptable shadow value.

If I slide the shadow edge control over to the left until I see the dots, (thresholding)... the correction is too far gone for shadow details.  When I bring the threshold over to the edge of the histogram (Approx. Value) I can get a good shadow setting but the black dots are not showing on the screen correctly.

It was confusing at first because I was going for a quick and dirty correction for proofing this image. I just about dropped a jaw bone when I saw what the thresholding did. 
These all seem to be pointing at the curve control don't they??

Do I have a bad version of it??
Greg

-default
Posts: 1916
Joined: Thu Mar 26, 2015 1:53 am

Postby -default » Tue Nov 11, 2008 12:36 am

Hi Greg,

I'm staring a hole in this, and after a couple of images, I still don't see a bug.  It seems to me there just a small number of pixels that show up on the preview image as black specks, with the bulk of the dark histogram being the darker areas of the sweater.

Can you get a similar problem working with a black-white gradient?  Or maybe I just have a blind spot to your explanation, in which case my apologies - I'm trying!

ggroess
Posts: 5342
Joined: Wed May 24, 2006 2:15 am
Contact:

Postby ggroess » Tue Nov 11, 2008 4:16 am

Let me kick a few other images around maybe this one is really weird...
I'll test a few and see if this is still an issue...

Greg


Return to “Curvemeister 3 Beta”

Who is online

Users browsing this forum: No registered users and 41 guests