Let's compare the dynamic range of our cameras

If the histogram under the RAW tab were indeed a presentation of values in the raw file, it would never change shape.

So, what does that histogram actually plot?

agedbriar wrote in post #11432893
If the histogram under the RAW tab were indeed a presentation of values in the raw file, it would never change shape.

So, what does that histogram actually plot?

Yeah, that's the question we're trying to address -- maybe there is something between the 12-bit and 14-bit cameras, because there seems to be a slight "nudge" to the right for 14-bit sensors, but it doesn't seem to be a cut-and-dry thing, or does it?

agedbriar wrote in post #11432893
If the histogram under the RAW tab were indeed a presentation of values in the raw file, it would never change shape.

So, what does that histogram actually plot?

IMO, it is the RAW data after application of WB (which changes the shape of the histogram) and before the application of Picture Style/TRC.

I would submit Eli, that every alteration that you make to that RAW image alters the histogram in some way.
I rely on it in LR to show and avoid burnout...the same with PS.

chauncey wrote in post #11433065
I would submit Eli, that every alteration that you make to that RAW image alters the histogram in some way.
I rely on it in LR to show and avoid burnout...the same with PS.

Chauncey - We are talking about DPP and its histogram on the RAW tab, which is not predictive like the RGB histogram (or the histogram in LR), or at least not to the same extent. Changing WB changes the shape of the histogram and Brightness (Exposure) shifts it linearly (without a change in shape) to the left or to the right, but the other sliders on the tab do not affect it (although Contrast, Highlight and Shadow change the S-curve superimposed on it).

yb98 wrote in post #11425929
I have noticed that the DPP RAW histogram gives information about the dynamic range of a camera.

That histogram does not provide any useful information about camera's DR for two reasons:

1. That histogram is the result of processsing the RAW file (that includes white balance, exposure corrections, tone curves,...), so the vertical lines don't represent EV stops.
   
2. Even if you plotted the RAW histogram in EV stops (which is totally possible, see here genuine RAW histograms in EV stops), you won't be seeing there the DR of the camera but the DR of the scene. Just counting how many 1EV-width columns contain data we can conclude a number of stops of DR for that scene.

Point taken Eli, I have never explored DPP. But I wonder about the practical value of this discussion.

chauncey wrote in post #11433286
Point taken Eli, I have never explored DPP. But I wonder about the practical value of this discussion.

PRACTICAL VALUE??? We're just flappin' our jaws.

Elie,

Shifting the histogram to the right by means of brightness boost actually does produce a change in shape, which happens in an interesting way. In fact, values near the right border (not just the furthermost value!) grow in amount progresively, not in a sharp, clipping step.

Let's keep the flappin' going

We are simply trying to find out what a DPP histogram, that one can evaluate and even work on, is actually plotting, as it doesn't seem to be very obvious, yet it would be useful to know.

_GUI_ wrote in post #11433244
you won't be seing there the DR of the camera but the DR of the scene.

That would be my guess as well...

We are just trying to understand the meaning of this raw histogram.
We are not saying that the RAW histogram data allow to know the DR.
What we are saying is that we have noticed that the white surface of the histogram on which this data can appear seems to differ from one camera to another. The fact that some part of the histogram are grayed indicates that no information can be recorded there, ok ? And the fact that some part of the histogram is white indicates that data can be recorded there, ok ?
So if this white part change depending on the camera model, doesn't this mean that some camera can record information where others can not ? Isn't this what is called DR ?

Originally Posted by _GUI_

HTTP response: NOT FOUND | MIME changed to 'text/html'

you won't be seing there the DR of the camera but the DR of the scene.
That would be my guess as well...

To nit-pick a little, would it not be an representation of that scene's dynamic range as interpreted by the camera's sensor which would not necessarily compare with what we visualize in our mind.
We use Photoshop/darkroom, or the like, to help us reconcile the differences.

Well, interesting question. I admit I haven't read all the documentation for DPP. I also wonder if there's a White Paper out there for this? There has to be a reason why the different cameras show different characteristics.

For those who don't "get" this discussion:

This is not so much a technical analysis of the characteristics about the cameras per se, but a discussion of the nature of the DPP histogram and how it may or may not relate to the individual cameras. What we see is a slight difference in the white point "hard stop" between camera bodies, represented by the "gray zone" to the right of the histogram.

I don't see that changing from scene to scene. It doesn't "appear" to change with camera settings, etc. In other words, it seems like DPP has the camera "coded in" to give that slightly different value.

So, the question would be, does that have something to do with the dynamic range of the sensor that DPP "knows", or is it just a fluke of the programming? Is it just a coincidence that the variation "seems" to give more of an "edge" in the highlights to cameras that we know to have a bit more performance in the high ISOs?

If you look at the histograms we've posted it's not huge, and who knows how much practical use it is, and if you don't use DPP you won't even have a clue as to what we are talking about, but still interesting.

And, really the only slider that moves the actual histogram around in the "white zone" seems to be Brightness, and it does it in a virtually linear way. The little peaks and vallies seem to be affected very little. The other sliders "seem" to only affect the little curve. And White Balance and Picture Style each have very little affect on the histogram, but Picture Style tweaks the curve a bit from what I can see. Of Course the RGB histogram will be affected more by these things.

I have found this pic on the web. I have never seen that before. There is no gray area at the left of the RAW histogram at all !
I don't know which camera is this and which version of DPP it is...

http://www.flickr.com …gner/2489881657​/lightbox/