Exposure is not how dark or how light an image looks!

ISO is just the relationship between exposure* and image lightness with the aim of outputting an average 18% grey image.

For me, the ISO setting is irrelevant when setting the exposure* as long as important highlights are not clipped.

I set the smallest f-number that will give my DOF and the slowest shutter speed that will meet my blur requirements without any consideration of ISO at all as long as important highlights are not clipped.

I don't care at all what setting the camera sets ISO to with Auto ISO as long as important highlights are not clipped because I set the final image lightness in post.

* exposure - amount of light striking the sensor per unit area while the shutter is open
** optimal exposure - the maximum exposure* within dof and motion blur requirements without clipping important highlights.
*** under exposed - more exposure* could have been added with the DOF and blur constraints still being met without clipping important highlights.

It looks to me like it's time for me to just move on, because "precision" and "photon noise" are not topics addressed or mentioned in the article for reasons mentioned numerous times. Best wishes to you.

He talks about precision, even if he doesn't use the word. And the fact that he left out the important thing while talking about the unimportant one is my point.

Here are some numbers that show how much noise there is in the upper tones of a modern camera:

So the 14 bit ADC becomes roughly an 8 bit one, for purposes of measuring the signal.

I see ISO as the inverse of EC. That is, I have an exposure value in mind and if the metering on my camera isn't getting me there then I adjust the EC -- which all happens before the exposure. I also have a brightness of the subject and/or scene in mind, and if the selected exposure doesn't get me there than I adjust the ISO -- which happens after the exposure.

I propose that it would be appropriate to lay this out to beginners in chronological order:

1. Metering
2. EC
3. EV
4. ISO

You've inserted an assortment of words and topics into discussing the article which are not found in the article itself because it has a narrower focus on other specific topics. The article doesn't use the word precision because, in my reading and perception of it, the focus of the article is: the method by which levels are distributed in linear unprocessed raw files; the effect exposure has on the distribution of levels in the linear raw file; and the redistribution of levels which occurs after nonlinear gamma correction is applied — not the ADC process by which those levels were derived, not their precision, not any sources or character of noise the levels may contain — but how and where those levels are distributed following exposure is the limited scope of this article.

We appear to have arrived at an impasse. Hopefully you feel that you've made your points. I feel the article contains concise and worthwhile information in an abbreviated form. Readers can decide for themselves.

If you want to discuss or make some additional points regarding ADC precision, noise sources and character, or any of the other numerous related and worthwhile topics which the article does not address; I would encourage it. But since they are not topics contained within the article I linked, I'd prefer not to continue down this same path of discussion and to move on to other topics. Thanks

To shoot welding, I use NIT sensor which has logarithmic response. Distribution of levels doesn't matter at all, all one needs is to assign the correct gamma.

take the 200 asa film out of the camera and how many images can you expose [😜

I have never photographed with film and so am referring to digital photography and I see you are unable to post anything that shows what I posted is not correct 🙂

And I also see you are still hiding from posting in your own words what the word exposure means when you use it in sentences.

he is absolutely right. how many images did he take without the film or glass plate in the camera using just shutter speed and aperture 🙄 i do it all the time testing a film camera 😁 works perfect for me.

thanks for your post, i was out all day and the same conclusion popped into my brain concerning the capacitors. now the big question for you Jim as my knowledge is a bit light on here. but doesnt the iso setting control the amount/level of electrical charge the capacitors can hold ? im most interested.

Every keen photographer sooner or later comes across and appreciates the mentioned papers. For the record, there are conceptual issues with the one above, which makes it misleading to the beginner. For example where the author says:

This is a major conceptual mistake. It gives the false impression that the input to the HVS should be non linear (e.g. gamma encoded) in order to make the image look properly bright. Nothing is further from the truth: cones in the retina also 'count photons in a linear fashion' similarly to what pixels do. To drive the point home he follows that statement with a dark looking image subtitled

These generic statements are wrong, or at the very least highly misleading. With a wholly linear workflow, how the final image will look depends on the characteristics of the output medium. If the output medium is linear, the displayed image should look right.

The fact that every beginner should be aware of is that light in photography behaves linearly all the way to our eyes - and ideally so it should whether it came looking out a window at a beautiful scene, or whether that window were replaced by a similarly sized monitor displaying a capture of the scene.

The HVS expects a linear light stimulus from the eyes. It then proceeds to process it non linearly, as it always does.

Jack
PS There can be benefits to gamma encoding, but those are not it. Today they have mainly to do with lossy compression.

i may as well change history. My theory is this. the recording medium in film was crystals (not important what they are made of) and they were exposed to light from a lens even a pin hole with no lens. now a digital sensor is a different recording device and myself and Jim have concluded that the recording medium for digital sensor is the capacitors that store the electrical energy from the photodiodes. my argument is that iso is part of the exposure process because its controls the amount of energy reaching those capacitors (recording medium) , in fact from my understanding the electrical energy from the photodiodes reach the iso control before the capacitors that store the energy,

conclusion: is that if the capacitors are the recording medium then iso is part of the exposure process. who would like to work out in Pico seconds if my theory is correct 😉 it will be a tight race but hey this is life changing 🤦‍♂️

ISO doesn't control the amount of energy reaching the capacitors.

That is not even close to being true.

Think of it this way.

For a given scene and lighting you set manual mode with f/8, 1/200s, ISO 100

After reviewing the image in camera you feel the image is still too dark but you cannot open the aperture or slow the shutter because your DOF and/or blur constraints will be messed up.

The only other option, assuming more light cannot be added to the scene, is raise ISO. You now set ISO 200 and the image looks good.

Both the ISO 100 and ISO 200 shots will have received the same exposure* (energy, to use your word) and so ISO had no affect at all on the exposure* (energy reaching the recording medium, to use your words).

* exposure - amount of light striking the sensor per unit area while the shutter is open
** optimal exposure - the maximum exposure* within dof and motion blur requirements without clipping important highlights.
*** under exposed - more exposure* could have been added with the DOF and blur constraints still being met without clipping important highlights.

But that's not ISO. ISO setting isn't ISO. BTW, with the same sensor different cameras use different ISO setting trigger values for the second gain.