Skip to main content

⚠️ Please note that this topic or post has been archived. The information contained here may no longer be accurate or up-to-date. ⚠️

Layer adjustments not limited to masked area.

Comments

7 comments

  • SFA

    Ric,

    This is a software discussion forum rather than an image editing comparison site. As such 2Mb should probably be enough to work with. No one needs hundreds of 150MB "sample" files downloading when they are not expecting them.

    If size is important to provide evidence for the problem just upload the image to a file sharing service and provide a link.

    Sadly, in this problem case, we may need to see the evidence rather than ask for the problem to be described. It sounds like this is one example where a picture might speak a thousand words.

    0
  • Ric Cohn

    SFA, not asking for 100's of MB. Screen shots are 2 - 3mb. I think 10mb would be more reasonable. I've uploaded to many forums and none have limited to 2mb. Because of this it's been years since I've needed to use a file sharing service. Frankly I'm old and I've forgotten how, but I'll look it up.

    0
  • SFA

    I sympathize with your view on this but surely a 2MB jpg at a typical screen resolution should offer decent visible evidence? (Maybe not so much for 4k or 5k screens ....)

     

    You and I might use such capacities with due consideration for others but I would lay down money that some would not.

    0
  • Ric Cohn

    I hope I've succeeded in linking to 3 screen shots showing layer 3 (named "Sharpn") on and off. 1. Mask visible. 2. Mask not visible. 3. Layer turned off.

    You can see the layer should only affect the bird. If you compare the layer turned off and on you can see the entire image is affected. This is most obvious in the foreground/left snow.

    You can also see the adjustments for this layer are to Clarity and Structure.

    https://www.dropbox.com/s/8cnwoicno8irvmz/Screen%20Shot%202022-01-31%20at%201.25.27%20PM.png?dl=0

    https://www.dropbox.com/s/nuwi615lciqr9kr/Screen%20Shot%202022-01-31%20at%201.25.42%20PM.png?dl=0

    https://www.dropbox.com/s/f50c5sj9y3kkkoj/Screen%20Shot%202022-01-31%20at%201.26.02%20PM.png?dl=0

    0
  • SFA

    Hmm.

    Very odd.

    How did you create the mask? 

    Manual brushing?

    Copy a previous mask (with or without subsequent adjustments)?

    Add a 100% mask and then delete all but the bird?

    Something else?

    I'm baffled. Just wondering if answers to these questions might spark an idea.

    0
  • Ric Cohn

    First added the layer, then manual brushing and then adjustments. As I said, the first two layers work correctly. I'm hoping for an explanation and solution, but thinking it might be a bug. I did report to support but hoping for user help to solve.

    I did some more tests and think this might an isolated problem, but I still don't know how often it happens or why.

    I just did a test that I should have thought of. Changing other adjustments such as curves, exposure, shadow, and highlight do only affect the masked area! I'm pretty sure that Clarity and Structure should also work but seeing that "mask failure" is limited to these (in the adjustments I've looked at) helps explain why I've not seen this before and increases the odds this is a bug.

    I also just tried this on a second image and can't easily reproduce it. And then I deleted the offending layer and made a new layer and it seems to work OK. Is there a cache or something for the image I should delete? I tried "Regenate previews" but that didn't fix the problem. So the good news is this is likely to be rare. The bad news is if it happens users might not notice while they are working on an image but will have their images made worse.

     

    0
  • SFA

    Sound strange but my guess is a rare and one-time data glitch of the sort that cannot really be predicted and will often evade capture by disappearing before it can be successfully captured and analyzed.

    The software guys might be able to dream up some obscure ways by which such a result may arise but that would still leave the problem of trying to prove the theories if such glitches are rare and random and cannot be readily reproduced.

    If you were to go to a doctor and you a similar question about some sort of random illness type symptom they would probably suggest keeping it in mind for a week or two but if there were no further occurrences of the symptom just carry on and forget about it. If you see repeats and can find a way to partially predict them, make another appointment at the surgery!

    0

Post is closed for comments.