Why are my photos noisy - discussion

Should be OK.

Here's what you're leaving on the table:

blog.kasson.com/a7riii/a7riii-shadow-noise-isos-500-640-4-stop-push/

Thanks, Jim. It's probably obvious, but I haven't worked out the lesson from your post. Setting ISO 640 leads to less noise than shooting at ISO500 and pushing? For anything higher than 640, set 640 and push? What about the lower values of 200,400?

I read your other post on strategy. Does the A7rii behave the same as the Riii?

You get slightly less deep shadow noise at ISO 640, but it's not a big deal. I think your plan to keep it simple and keep the ISO low will get you the kind of images you want without the complexity that you don't appear to want. I tend to treat the camera as if it had two base ISO settings.

Yes, I got that from your strategy post.

If you expect trouble from highlight clipping, shoot a set of images with auto exposure bracketing. (I assume most cameras allow this.)

Don

Two base settings makes sense. ISO 100 for tripod or flash. ISO 800 for hand held and general walkabout.

Don

I also prefer to treat the camera as it has two base ISO settings. However, that approach can make the EVF too dark, which makes it harder to see details. At least with Nikons, a darker image in EVF also degrades AF.

It also depends on the light. If you can shoot handheld at ISO 100, you should do it.

Same with my Canon R5 and R7. If "exposure simulation" is enabled, and low light results in a dark image at max ISO, AF becomes hopeless. Turn off "exposure simulation" so that you get "auto-ISO" for the EVF, and the camera can easily focus.

It is very convenient if a camera supports a one-button switch for exposure simulation. DSLRs don't have that issue :).

I last took a physics class in 1968 and likely have forgotten much. I think I understood all that you wrote until

.
I might have guessed that a warmer sensor chip would produce photons from black body radiation and some of those photons would be far enough from the peak in energy level or frequency to produce an electron in a photosite. Is that what you are talking about, or something I do not understand?

Thanks in advance.

Maybe warm electronics just free electrons directly? It'd be mildly interesting to know.

I mean the effect of thermally generated electron-hole pairs.

Dark (leakage) current doubles every 10 degrees C or so, and needs no photons to generate the excess electrons resulting from the leakage.

An example is the Sigma fp, one of which I own. Nice camera except that it is not good at flash.
Don Cox

We all used to shoot ASA 125 film hand held. But I'm not sure the results are all critically sharp.

Don Cox

It's an Arrhenius law thing. There's an activation energy - or rather a range of activation energies, depending on the crystal defects near your pixel. As JK said, (at temperatures around 300K) the rate doubles roughly every 10 kelvin.

[edit:]

The rate is quite low, for most pixels, in most photographic conditions - or photography with these sensors would be a bit rubbish. The accumulation of "dark current" electrons has a Poisson distribution, just like regular photo-electrons. The Arrhenius law for "dark current" is why it's good to have a cold sensor when taking long exposures of dark things.

As you know, the loss of sharpness was not due to low ISO but to the required shutter speed.
Current cameras often have either IBIS or lens VR, so base ISO usage is more frequent (unless subject motion prevents it).