Actual raw histograms

https://www.rawdigger.com/ excerpt

General Tools and Features

  • RawDigger allows one to display and research the real Raw Histogram of the raw data, which greatly differs from the in-camera histogram, as well as histograms presented by most of the raw converters.
  • To facilitate accurate raw image study, the histogram can be displayed in different modes, over the whole image or rectangular selections of the portions of the image.

End excerpt. Are the displayed histograms in PL5E accurate and precise or do these have the limitations alluded to in the above excerpt? I am not referring to the various display and statistical analyses provided by RawDigger but merely if the histograms in PL5E distort the photographer’s view of the raw image pixel data (eg, exposure).

For what its worth. I opened the same raw image in Photoshop, Capture One, and Photolab and all three histograms were virtually identical. I do not have RawDigger so I have no idea if that application would have a different histogram. If there is any distortion of the raw image, all three application are interpreting the data in the same manner.

May one ask as to which raw format (body, specifications when the “raw” has more than one format as Nikon allows) you compared?

It’s an interesting question: my guess is that PL5 is displaying the raw histogram and not the histogram of the Jpeg ‘thumbnail’ often included in the raw file (and that is displayed on the view-screen of cameras as you compose and shoot!).

My guess is based, however, on eyeballing the Luminance histogram in PL and comparing it to the raw histogram in Fast Raw Viewer (+/- the same, but less detailed, that RawDigger displays).

There are small differences in the shape of each of the R, G & B curves in the histograms in PL and FRW. But I’m not sure if they are significant because the different scaling of the histograms makes a precise comparison difficult.

Certainly the changes that appear in the histogram in PL when you adjust the various development tools seem consistent with the changes you would expect in the raw data.

Assuming, that this is the raw data frequency distribution that is being displayed I really wish DXO would adopt the practice of showing the current luminance ‘histogram’ in the background of the Tone Curve control chart (as e.g. LR, C1 etc do). This makes a BIG difference, in my view, to the accurate use of that powerful tool.

I also wish they’s allow us to “pull out” the Tone Curve control as a floating window so it can be enlarged to a more useable size than the current, narrow, palette permits.

DPL’s histogram is based on the processed image. If it were based on raw image data, the histogram should not change, no matter what we do with exposure, contrast etc.

1 Like