Library Histogram Not Showing Clipped Whites; Develop Histogram Does Show Clipped Whites

  • 2
  • Problem
  • Updated 5 months ago
I have a photo that has some clipping in the whites. The develop-module histogram shows a short vertical line at the extreme right, properly showing the clipping. Switching to the Library module, the histogram is nearly identical, however it's missing the vertical line showing the clipping! It's quicker to cull photos in the Library, but now I can't trust the histogram, so I can't trust I haven't blown out some whites. The histograms should be identical.
Photo of Scott Rosenthal

Scott Rosenthal

  • 15 Posts
  • 0 Reply Likes

Posted 5 months ago

  • 2
Photo of Andrew Rodney

Andrew Rodney

  • 476 Posts
  • 76 Reply Likes
Neither histogram shows you a raw histogram. Any clipping scene is an attribute of the current rendering settings. While the two should match exactly, the most accurate indication of the data is always found within the develop module.
Photo of Scott Rosenthal

Scott Rosenthal

  • 15 Posts
  • 0 Reply Likes
True, but that's not what I'm referring to. These are images that have had no development done to them--straight out of camera. The histogram on the left is from the Develop module. The right one is from the Library module. They both should be showing the same blown-out whites.

 

Photo of Andrew Rodney

Andrew Rodney

  • 476 Posts
  • 76 Reply Likes
If raw, there's always some development. If JPEGs, that's what it is (I'd again judge in Develop). 
Photo of Scott Rosenthal

Scott Rosenthal

  • 15 Posts
  • 0 Reply Likes
If culling through images was faster in the Develop module, that would make sense. But what use is it to have an incorrect histogram displayed in the Library module? Or better, yet, what's the advantage to having incorrect information shown as true information?

To me, it seems as if this issue isn't about raw, JPEG, straight-out-of-camera, etc., it's about something so basic in the histogram as if the pixels are saturated or not. It's almost like the Develop module displays 256 bins (0 to 255) and the Library module displays 255 bins (0 to 254).
Photo of Johan Elzenga

Johan Elzenga, Champion

  • 1047 Posts
  • 401 Reply Likes
The histogram is based on the preview. Develop previews are in a special version of ProPhotoRGB, library previews are in AdobeRGB. Maybe that causes the difference you see, although one would not expect differences in clipping because of this. When I toggle between library and develop I often see slight differences between the histograms too.
(Edited)
Photo of Andrew Rodney

Andrew Rodney

  • 476 Posts
  • 76 Reply Likes
While the previews differ in color space as you point out, the Histogram's do not (or should not). According to Eric Chen at Adobe: The histogram always reflects rendered results and not the unrendered source image.

The differences above should be larger visually if the two color spaces were different. The clipping seen in one (more accurate) but not the other could/should be considered a bug? 

Probably the thing to do is take an image where we know the color gamut exceeds Adobe RGB (1998) and compare the two Histograms in both places. Also, I have seen differences toggling color spaces in ACR where clipping IS seen in sRGB but not ProPhoto RGB (shown in my video) however the clipping is clearly saturation clipping, not all three channels as we see here. 

I again believe the Develop module and Library should show identical Histograms but don't. And I again believe that the place to view this correctly is in Develop. Perhaps we need to requests: fix what appears to be a Histogram bug in Library and, a feature request to show us a true, raw Histogram as we see in RawDigger. 
Photo of Andrew Rodney

Andrew Rodney

  • 476 Posts
  • 76 Reply Likes
If you load this TIFF into LR, it becomes clear that Library is getting RGB/255 (Lstar 100) wrong
http://www.digitaldog.net/files/Gamut_Test_File_Flat.tif
There a lot of 'paper white' (Lstar 100) in this document and one should see this clipping. And one does in the Histogram in Develop, not Library. I believe Library is wrong. This should have nothing to do with color space, gamut etc. Even though this is a rendered TIFF, RGB255 in all channels should show a spike. But we don't see it in Library for whatever reason. 
In Develop, with the above image, setup a soft proof between the normal and Adobe RGB (1998) and the differences in the Histogram are clear (colors clip in Adobe RGB (1998) ) yet the spike remains in both. It's not a working space difference. It's a bug.
(Edited)
Photo of Scott Rosenthal

Scott Rosenthal

  • 15 Posts
  • 0 Reply Likes
With your TIF, notice that the spike in the zero bin (level) is there in the Develop module but not in the Library module. Both ends are missing.

Agreed, it is a bug, or error, or undocumented feature. 
Photo of Scott Rosenthal

Scott Rosenthal

  • 15 Posts
  • 0 Reply Likes
Agreed that one shouldn't see differences in the histograms with clipping, if the data is clipped.

I looked at the raw file in a different program, RawDigger, which reports that there is clipping in all four channels (R,B,G1,G2)

I also, used ACR and changed the color space to all the available ones, and the histogram always showed the clipped whites, which is what one should expect.

My take is that I can't use the Library histogram to cull photos with blown whites, if that is important to me. I'd really like to hear Adobe's take on this.
Photo of Andrew Rodney

Andrew Rodney

  • 476 Posts
  • 76 Reply Likes
You need to understand what the CURRENT Histogram is telling you! They lie too:

Everything you thought you wanted to know about Histograms

Another exhaustive 40 minute video examining:
What are histograms. In Photoshop, ACR, Lightroom.
Histograms: clipping color and tones, color spaces and color gamut.
Histogram and Photoshop’s Level’s command.
Histograms don’t tell us our images are good (examples).
Misconceptions about histograms. How they lie.
Histograms and Expose To The Right (ETTR).
Are histograms useful and if so, how?

Low rez (YouTube): https://www.youtube.com/watch?v=EjPsP4...
High rez: http://digitaldog.net/files/Histogram...
Photo of Andrew Rodney

Andrew Rodney

  • 476 Posts
  • 76 Reply Likes
What we need is a raw Histogram as you see in Rawdigger but none exists in Adobe products.