Re: ISO for beginners

Sorry .. I carried this conversation further than intended. I use to read many of these discussions back on DPR and found them interesting to the physicist and sensor designer, IMHO, when talking about noise, etc and to the optics and lens designer when discussing lens qualities... Many years ago I knew more about the Hydrogen atom than I would ever need to know .. Today my interest is in composition and noise when it intrudes on my images on a 27 inch monitor or on an 8X10 print. And about noise I only think about how and when to eliminate it or how to use it.... I want to think about these processes and how they impacts my camera operations as I take a picture and maybe how they relate to my PP nothing more ... So while interesting for me they would be more useful, again IMHO, if the poster thinks about the reader and how he can use this information … But then this is a photo/discussion forum so ….

WhyNot

Photon counts are related to both the intensity of the light reaching the sensor, and the shutter duration.

Consider two shots of the same static subject. One at f/2 and 1/1000, the other at f/5.6 and 1/125. At f/5.6 the light reaching the sensor is 1/8 as strong as the light reaching the sensor at f/2. However we are keeping the shutter open 8 times longer, to the photon counts are essentially the same.

At the same ISO setting, both captures would produce a JPEG with the same lightness, even though the f/5.6 capture had a lower intensity of light reaching the sensor. If we only cared about the intensity of light reaching the sensor, then shutter speed would be irrelevant to image lightness.

Think of photons as raindrops of light, and pixels as being buckets catching those raindrops. Light intensity is how many raindrops are falling per second. The number of raindrops in a bucket depends both on the intensity of the rain, and how long the shutter is open. Even if it only a very light rain, we can catch a lot of raindrops by keeping the shutter open for a long time.

I think the raindrop analogy is a good way to explain to beginners. People understand that a bucket left out in the rain captures more raindrops if the intensity of the rain is higher, or if you leave it out longer. Very roughly speaking, pixels and photons work the same way. We capture more photons when the light intensity on the sensor is higher, or the shutter is left open for a longer period of time.

No complex math is needed, and most people are familiar with rain and buckets.

You are so kind and considerate to think about me and my brain.
Thank you !!!

( I was worried to have to count photons using my fingers - before shooting and setting the camera )

😂

👀 🙄 🤷🏿‍♀️ 🔜

OT

Icon's draught solved

I believe that raindrops falling in a bucket also have Poisson statistics. And the high conversion gain setting could be modeled by a funnel the same diameter as the base ISO bucket's diameter feeding a cylinder as high as the base ISO bucket but smaller in diameter.

I just watched a Richard Feynman lecture on quantum electrodynamics where he used the raindrop analogy for photon arrival. You are in good company.

ISO is also a setting of our cameras; sometime referred to as “ISO speed,” but most often as “ISO.” In fact, almost everyone in this thread uses it with that meaning.

The resulting file in my case is a RAW file. It has no lightness. It does have an embedded JPEG but the same lightness can be achieved with various mappings from photons to RAW values. I would say that ISO is the mapping from photons to RAW values, and one may want to include the way this is done. It depends on the camera, like it or not. As such, it is baked in the RAW. Modes like highlight priority change that however.

with my limited knowledge of sensors the only thing i can compare it to, is the RF gain control on a sw radio, not the audio volume control. RF gain control controls the the amount of RF signal strength from the antenna . if you have very hi gain antenna the RF signal can be to strong and create distortion, its just like the ISO control on a sensor not the levels control in say photoshop.

There are cameras where the ISO setting does not affect the raw data. With these ISO only changes how the data gets interpreted.

Modes like Highlight Tone Priority generally don't change the values. They only alter how those values are interpreted.

These are implementation details, that vary between cameras models.

Conceptually, raw data reflects a count of photons seen by each pixel. These counts may be scaled, either by an analog or digital process. That scaling may be altered by the ISO setting. Thus the context of the specific camera model is also important when processing the raw data. Keep in mind that at the same exposure, cameras with bigger pixels will see more photons per pixel.

For a beginner audience, I think it is best to stick to a conceptual framework, and not try to promote a particular implementation. Nor do I think it is necessary for a beginner to understand the specific mechanisms used by their camera.

The RF gain control affects a circuit that takes voltage in, and yields a different voltage out. In that context "gain" makes sense as both the input and the output is the same thing (electrical voltage).

ISO is a very different beast. The input is "photon counts per unit area" and the output is "image lightness". Note that ISO is not a mapping from brightness on the sensor to brightness in the final image. Given the same subject, f/2 will project an image on the sensor that is eight times brighter than at f/5.6. Yet at the same ISO, if the f/2 shutter speed is 1/1000 and the f/5.6 shutter speed is 1/125, we get the same results in terms of image lightness. Keep in mind that longer shutter speeds do not make the image on the sensor "brighter", but the sensors do see more photons.

Think of ISO as being like the curve for grading a school test. The curve maps the number of correct answers to a letter grade. Depending on the curve, twenty correct answers might produce a "C" or an "A". I don't think the grading curve should be called "gain", as the input (number of correct answers) is a different thing than the output (a letter grade).

I did not say that it must change the RAW data. You just need to know how light and RAW values are related.

When I was a beginner in the digital photography, I wanted to understand exactly what I am now talking about.

What happens after the ADC is more or less like the audio volume control.
In some cameras / for some ISO settings digital multiplication is used to reach the ISO goal, and it can be also left to a raw converter, in full or in part.
We have something akin antenna gain too.
ISO "implementation" isn't necessarily a one-step deal, it can be (and often is) a rather multi-stage ladder.

I'm strongly in favour of this.

I guess my original thoughts were based on the audience you want for this discussion. On DPR and now here there are two distinct audiences –what I would refer to a gear oriented and those that are picture oriented. The gear oriented is interested in the beauty and limitations of the technology that went into their gear. And want to understand it. The picture oriented wants to know the simplest way to think about the settings he/she has available to him and perhaps the consequences of that decision for the file he'll get to pp and display … as he looks at a scene and has seconds or minutes to make that decision as it or the light will change … The question is really about where and how you adjust the histogram if you are looking at that on the EVF or screen and what are the consequences of that decision .. Buckets, wells, pans its all the same -- collectors of different depth and I'm not sure anyone from the picture audience wants or needs to discus waves vs particles or De Broglie …. I know I am no longer need that but occasionally enjoy the resulting arguments – the what if? But then …

WhyNot

BTW I'm not sure this reply is a good idea but fragments of my other entries keep showing up here .. and I have no intention of returning to make another reply and apologize now for this one ....