Pages: [1]   Go Down

Author Topic: What is Lightroom doing when I "set the white point" with the white slider  (Read 1053 times)

alecdann

  • Newbie
  • *
  • Offline Offline
  • Posts: 4

What is Lightroom doing when I shift+double-click the white slider to "set the white point"?  I assume LR is trying to find the optimum setting to minimize white clipping, but when I move the slider I see changes in other parts of the histogram so I wonder what is going on across the tonal gradient.  I'd like to be able to explain this to a student.

(Note: I put "set white point" in quotes because I think it's a confusing use of the term.  Setting a white point in monitor calibration means setting the color temperature of white in the monitor.)
Logged

BradSmith

  • Sr. Member
  • ****
  • Offline Offline
  • Posts: 566

I'm on LR 5 and shift double click on the white slider in the Develop Module does nothing.  What version are you using
Logged

alecdann

  • Newbie
  • *
  • Offline Offline
  • Posts: 4

Lightroom CC 2015.10.  Not sure which release this feature was introduced but it has been around for a few years.  This 2014 article by Scott Kelby explains how to do it: http://lightroomkillertips.com/lightroom-auto-set-white-black-points/
« Last Edit: April 25, 2017, 11:40:55 AM by alecdann »
Logged

BradSmith

  • Sr. Member
  • ****
  • Offline Offline
  • Posts: 566

Thanks, I went back and did it on a different image and it functioned as you said.  My first random image choice, which I'd previously adjusted, may have been so close to the outcome the software choose with the Shift double click, that I didn't notice any difference.
Brad
Logged

luxborealis

  • Sr. Member
  • ****
  • Offline Offline
  • Posts: 2376
    • luxBorealis.com - photography by Terry McDonald

From my limited understanding of the under the hood stuff, I believe LR is simply "looking" for the highest values pixel(s) and setting it (them) to 100%-100%-100%. At times, though, I notice auto White Point is set to a negative number, which indicates to me that my blown out pixels are being re-mapped downwards, although "down to where" I'm not exactly sure as they would be at pure white if they were blown out.

If LR is being true to the tone mapping idea, then, I am also surprised that shifting the White Point value has effects beyond pixels from that upper range.

In photos I know should have the full range from pure white to black, though, I am having excellent success with using auto White and Auto Black Point in assisting with determining those base values.
Logged
Terry McDonald - luxBorealis.com.
Flickr Account;
PhotoBlog Read and subscribe!

alecdann

  • Newbie
  • *
  • Offline Offline
  • Posts: 4

Terry,

I think you are close to the answer.  I've noticed the same thing about the white slider going to negative, however, and have attached and example where the white slider goes to negative (these are zoomed to 1:1 and show unadjusted and after shift-clicking the white and black sliders).  The reduction in the white slider helps reduce the clipping but not all of it.  The black slider is also reduced, however, and the adjustment introduces clipping. 

What's most interesting to me is that the midtones show more separation and I've seen it often enough that I'd like to really understand what's going on.

If I had to guess it's that LR is resetting the tonal range to optimize the image in general.  But as Stevie Wonder says, "when you believe in things that you don't understand, then you suffer."



Logged

Simon Garrett

  • Sr. Member
  • ****
  • Offline Offline
  • Posts: 729

From my limited understanding of the under the hood stuff, I believe LR is simply "looking" for the highest values pixel(s) and setting it (them) to 100%-100%-100%. At times, though, I notice auto White Point is set to a negative number, which indicates to me that my blown out pixels are being re-mapped downwards, although "down to where" I'm not exactly sure as they would be at pure white if they were blown out.

If LR is being true to the tone mapping idea, then, I am also surprised that shifting the White Point value has effects beyond pixels from that upper range.

In photos I know should have the full range from pure white to black, though, I am having excellent success with using auto White and Auto Black Point in assisting with determining those base values.

From my observation, setting auto white point (hold shift and double-click the white slider label) sets the range so the brightest pixel is 100% (and similarly for black slider so the darkest pixel is 0%).  However if you use auto tone (click the "Auto" button) then it can push the white much higher than this.  It seems to use a different algorithm where it decides what are blown or specular highlights, and lets them get completely blown out.  If you then do auto-white, that may result in negative values for the white slider (as the auto on exposure may have taken the whites very high). 

I'm not sure if my explanation corresponds to what LR is trying to do, but that seems to be the effect. 
Logged

alecdann

  • Newbie
  • *
  • Offline Offline
  • Posts: 4

Thanks, Simon.  Your answer makes sense, though I expect there's something a bit more complex going on.

I posted this question on the Adobe Lightroom forum as well and got a couple of answers from Jeff Schewe. 

He first responded that "The -20 Blacks that result in a blue channel clip is actually to be expected. And while it's correct that the Whites and the Blacks settings "manage clipping" that's not to say you should NEVER clip, just be aware of it :~)  Sometimes intentional clipping helps an image..."

I then asked what Lightroom is optimizing for and he answered: "It's optimizing the black point according to Thomas Knoll's opinion of what a black point should be. It's not always 100% correct, more like 80/20 and sometimes does something a bit less than optimal, but it usually is in the correct direction. From the result, season to your own personal taste :~)"

Here's a link to the full thread: https://forums.adobe.com/message/9478134#9478134

I've attached screen shots of the image he's referencing (before and after screen shots, zoomed in to show the areas where clipping is indicated -- these are the shots I forgot to attach in an earlier post).

I think I've gotten about as far as I need to on this.  I'll keep experimenting with it because what's happening in the mid-tones is interesting.

I appreciate your help.
Logged

Simon Garrett

  • Sr. Member
  • ****
  • Offline Offline
  • Posts: 729

Thanks for the update from Jeff.
Logged
Pages: [1]   Go Up