Re: ISO for beginners

I believe that that is an unnecessary dichotomy, and one that does not describe the totality of the situation. I believe that the better one understands their tools, the better they can employ them to make good images. I don't think I am alone in that belief.

And what if it is a wrong question? ;)

To add some fuel in this unnecessarily complicated discussion, no, the amount of rain drops does also depend on the size of the bucket.

This thread nicely illustrates the need for an FAQ area that features short, accessible posts for new photography enthusiasts about core topics.

Yes, and the pixel size is a factor in the number of photons reaching a pixel.

In order to know the rate of rainfall, one must know the bucket size, the number of drops it captured, and how long it was sampling the rain.

But we have strayed far into the weeds. A beginner doesn't need to know the implementation details. All the beginners needs to know is that ISO gives the context for mapping the photons captured by the pixel to lightness in the JPEG. A beginner doesn't know the breakdown as to which parts are implemented in analog hardware, digital hardware, or the software that processes the data from the raw file.

That's easy. The raw values are essentially counts of the number of photons that reached that pixel while the shutter was open. These counts may be scaled in some way. As more photons reach the pixel, you get higher raw values. In the vast majority of cameras, each pixel is behind a colored filter, so you can't draw a definitive conclusion from a single pixel as to how dark or light the scene was at that point.

I know a lot of beginners in digital photography who skipped shooting JPEGs.
I think we are not here to decide what beginners need to know, but to answer their questions with facts.

When I was a beginner in digital photography, I shot raw, or used Kodak PCDs, or PMT scanners. The only time I shoot JPEGs is for testing or when I'm using a cellphone.

Is it factually incorrect to say "The raw values are essentially counts of the number of photons that reached that pixel while the shutter was open. These counts may be scaled in some way. As more photons reach the pixel, you get higher raw values. In the vast majority of cameras, each pixel is behind a colored filter, so you can't draw a definitive conclusion from a single pixel as to how dark or light the scene was at that point."

If it is not factually correct, how would you suggest correcting it?

My goal was not to give a detailed overview of the various ways ISO can be implemented, but to give a conceptual overview of how ISO affects the mapping from photons hitting the sensor to image lightness in the JPEG.

I am more than happy to see better overviews.

Think of it a bit, because, yes, it's incorrect.
Pointing every beginner to JPEGs is wrong. Presuming what "All the beginners needs to know" is odd.

The ISO Speed standard does not actually apply to raw files.

The standard considers processing from photons on the sensor to image lightness to be a "black box". The inputs and outputs are specified, but not the internal mechanisms. As the raw file is something the standard considers to be "inside" that black box, it is an "implementation detail" not addressed by the standard.

For those shooting raw, the ISO setting on the camera may not be the actual ISO speed used when processing the raw data.

Personally, I think that discussing how ISO affects camera produced JPEGs is a reasonable place to introduce ISO.

The raw values are scaled counts of the number of photons to which the pixel responded. Photons that did not produce electrons are not counted. I think is is unnecessary to say the counts may be scaled. There may be cameras that have an ISO setting that is the unity gain ISO, but I've never seen one.

and there is clipping. Raw DNs give no account for how bright the scene was, too. One can shoot the same scene under the same lights and arrive with different raw DNs, even if the exposure is the same.

Yes. To the extent there are "two distinct audiences" who are on either site of the perceived dichotomy, perhaps the members of both audiences are missing half the picture (to coin a phrase).

However I believe there is a third distinct audience - made up of the people who recognise the interdependence between tools and techniques on the one hand and results and artistry on the other. We should (IMO) be encouraging folk who are engaged in disputation about what is wood and what is trees to take a step back and observe that it is a forest.

Yes, there are boundary conditions. There is an upper limit to the number of photons that can meaningfully be reported. Furthermore, the numbers may not be 100% accurate. We both agree that the raw values are not a direct reflection of how bright the scene was at that point.

Hence the use of the word "essentially". This makes it clear that the statement is about the fundamental nature of the relationship, and makes it clear that it does not encompass every possible detail nor every possible situation.

So let me repeat my question. What is factually wrong with "The raw values are essentially counts of the number of photons that reached that pixel while the shutter was open. These counts may be scaled in some way. As more photons reach the pixel, you get higher raw values. In the vast majority of cameras, each pixel is behind a colored filter, so you can't draw a definitive conclusion from a single pixel as to how dark or light the scene was at that point."?

Yes. Pixels are not perfect photon counters. Furthermore, most cameras have filters in front of the pixels that prevent many pixels from ever reaching the sensor. However, for an overview, I think it's fair to suggest that pixels can be thought of as photon counters. Yes, they are imperfect counters, but essentially they count photons.