Raw ETTR example case

An ETTR shot with a Hasselblad X2D developed with Ps default settings:

The raw histogram:

Proof that it's ETTR is this image with 1/3 stop more exposure:

Here's the back of the camera histogram:

That makes it look like the highlights are blown. In fact, you have to underexpose 1 2/3 stops from that image to get one where the in-camera histogram doesn't make the image look blown.

If you bring the image into Ps with the default settings, it looks fine:

Here's a Lr histogram of a shot a stop underexposed from that exposure:

Clearly, Lightroom is applying a shoulder to the highlights in the ETTR image.

the raw histogram is blown in the green channels in the second image. does your camera have an exposure fine tuning feature like sony ?

That's the point. That is the image with 1/3 stop more exposure.

Isn't that the point? The second image has 1/3rd of a stop more exposure and is clipped. The first image is not. This proves that the first image is as ETTR as you can get.

Oops, typing too slow. Repeating your point 🙂

We are in violent agreement.

LrC histograms also depend on the profile applied. For the same picture, the histogram will show clipping with some profiles and no clipping with others.

Yup. Because this was with the default Lr settings, the profile was Camera Standard. That is the most accurate Lr profile I know of for the X2D.

I have noticed that if you take the same image into fast raw viewer it does not exhibit the shoulder that ACR has. so that makes me think that ACR is applying some camera settings and the fast raw viewer is not.

ACR, like Lr, applies a tone curve that is dependent on the profile. FRV doesn't do that to raw files, because then it wouldn't be a raw histogram.

Here's the Phocus histogram with default settings:

But there are highlight warnings, even though the raw is not blown:

I have a question Jim, like sony my A74 is advertised DR up to 15 stops and so does your Hasy, why is the raw histogram only showing 10.5 when clearly the in-camera and other histograms are edge to edge ?

Those 15 stop dynamic range numbers are marketing bushwah. Downhill, with a tall wind. Look at Bill Claff’s PDR numbers, which make sense for smallish prints.

The histogram xaxes are different.

Something I find annoying about RawDigger is that there isn't an option to have it set EV0 to the clipping point (which it usually knows).

When posting graphs like this, I think it's better to manually set EV0 to the clipping point.

Also, log scale on the y-axis makes it clearer how many pixels are clipped, and the recorded dynamic range.

The clipping point is pretty obvious if you look at the second raw histogram. I thought it would be confusing for some people if I mixed the y axes in the plots.

Sorry, I cross-posted an edit of my post with your post.

The 3EV down from clipping, rounded to a power of two thing can, and has, created some confusion. I personally don't care what a light-meter might say are the mid-tones - I just want to know where RAW clipping is. Which is something Michael Reichmann was asking the manufacturers for 20 years ago.
I'll decide on the mid-tones.

Those 15 stops have a footnote saying "Sony test conditions," which nobody knows about.

I'm with you on that point.

Here's what I'm guessing. They measure EDR with a SNR threshold of 0, then scale to a print size that gives them the number they're looking for. And while they're doing that, they barricade the developers in the lab, or maybe muzzle them.

its totally over the top , i was wondering if they lift the shadows 3 stops and add that on to the actual DR 🤨

Hi Jim, i just took 2 images 1 in camera ev0 and another ev+.7 took the raws into ACR and increased ev image 1 by 1/3 stop and decreased image 2 buy 1/3 stop. here are the 2 histograms . i found little variation.