ISO & Dynamic range - what is going on?

In a camera, the ISO setting normally affects the gain between the sensor output and the raw written data. In some cameras, that gain is fixed and the ISO gain is applied in the conversion software. In the green curve, the gain is proportional to the ISO value; in the other curves, it is not. The blue curve (your Fuji) looks to me like it has "extended" low ISO and high ISO; and the black curve has some trickery in addition to that.

"Extended" means that, for example, that the camera aperture and shutter values are not changed by the ISO setting below "native", say 100 and some high value..

The graphing site owner Bill Claff does respond to questions and can give a much more specific response.

I dont take any of these testing sites (linked to manufacturers 🫣) for real any more, i borrowed a sony a7r5 the other week and put it through some simple tests. it doesnt compare to the a7iv for DR and hi iso performance , just as i had thought, and even worst case is the a6600 compared to the a6700. the a6700 is so far ahead its nearly FF performance based on images i looked at from past cameras with the same sensor, baseing DR on noise is just plain wrong, as a guy said on FM forums how can increasing noise reduction(reducing noise) increases DR 🤔its all a joke. just do your own testing thats the only way you will get the answers.

quote from Apples new 20 stop sensor " Patent quote : "advanced on-chip noise suppression "

ymcinema.com/2025/06/25/apple-just-patented-an-image-sensor-with-20-stops-of-dynamic-range/

www.dpreview.com/forums/post/68332406

Basing DR on noise is just correct.
quote from AI:
Relationship between Dynamic Range and Noise Floor:
-A low noise floor is crucial for achieving a wide dynamic range.
-A high dynamic range requires a sensor that can capture both very bright and very faint signals without being overwhelmed by noise.
-Noise reduction techniques can improve both dynamic range and perceived image quality by lowering the noise floor.

The graphing site owner Bill Claff does respond to questions and can give a much more specific response.

The ISO definition of DR is based on the noise floor specified as the point when the signal-to-noise ratio is 0 EV.

Discussed by a human here:

dougkerr.net/Pumpkin/articles/ISO_Dynamic_range.pdf

the opening quote says it all

" The dynamic range of a digital camera can be simplistically defined as the ratio of
the maximum and minimum luminance that a camera can “capture” in a single
exposure."

its pretty funny that isnt it, so last night i was looking at some images from my past a6300 ( same sensor a6600) taken in my studio and instantly noticed the
higher noise at the edges of my studio white backdrop. there is no noise in my a6700 images yet the a6600 reviews show higher DR 🤨🤣 facts are more important to me than donations to review sites. 😁

Sorry, but the opening quote does not "say it all". On the other hand, International Standard ISO 15739 tells us exactly what those two luminances are and exactly how they are measured and exactly what formulae are used to calculate the ISO DR value.

so shooting the a7r5 a6700 and a7iv with the identical extreeme macro machine where nothing changes and all settings remain the same the a7r5 images were all noticeably darker 🤨 what does that tell you 🤔 and the reveiw test sites all say the a7r5 has the highest DR 🙄its so obvious what the "marketing effect" has over the numbers. there we go i have just invented a new equation to the camera industry 😎