I may be misunderstanding things here. My camera (and previous ones) has a setting that allows you to go below the 'minumum' and above the 'maximum' iso. But... I understood that changing the ISO just increased or lowered an electrical charge somewhere (let's not get into too much detail here) in the system so that the end effect was to raise or lower the brightness (or whatever you call it) when the data is stored.
So, if that is true (and maybe it isn't), why do we need to have a 'special' setting? For example - my Fuji X-T5 goes up to 12,800. But by setting something, it can go up to 25600 or 51200 ISO. So, why not just let it do that? Why limit it?
To my (limited) understanding the ‘electrical’ changes extend only over part of the range. Over and under that, changes are digital. At some point, images at that setting do no longer completely meet the quality level the camera maker sets. There are several points where the camera maker can make choices applying the standard, this is one
Also to my limited understanding - native ISO is the range that the sensor and allied circuitry covers - extended ISO requires further software processing to simulate higher/lower ISOs
When I read our "tame Gurus" discussing this sort of thing, I think I kind of understand it - but by the next day, the details have disappeared from my little brain! So, I guess we'll have to wait for one of them to happen along for the definitive answer
Look at Bill Claff's Photographic Dynamic Range Chart: www.photonstophotos.net/Charts/PDR.htm#FujiFilm%20X-T5
He says "triangle down indicates noise reduction" and "Open symbols indicate values outside the normal analog range".
I think it means "digital".
X-T5 seems to have it's base ISO at 125. Lower values are outside the normal range, but still very usable, I guess. I have used quite often ISO 50 in my Sonys with good results. But never used ISOs higher than 12800.
BTW. You can compare cameras selecting them in list at the right.
There are no hard rules, but at the high end, it is probably a way for the manufacturer to say "I told you so" when you enable the extra high ISO range, when you complain that there is too much noise. Most cameras also don't increase analog gain for this range, but that is not the reason for most of the noise. There is no consistency here, either, with what the highest "extended" ISO is. The Nikon D5 goes up to ISO 3.2 million, but the Canon 5Dsr only goes to 12,800, which makes no consistent sense, since the D5 has way more noise at ISO 3.2M than the 5Dsr does at ISO 12800.
At the low end of the ISO range, extended low ISOs may have less headroom relative to middle grey for that ISO, or the highest raw highlights are non-linear, which can cause color shifts if the software does not correct it.
If there was less headroom for the highlights in the extended low ISOs, there should be less PDR, I guess. Not visible at ISO50.
But what is happening at very high ISOs? The charts are very similar at the low ISOs but very different at the high ISOs with X-T5 and my 7R4. www.photonstophotos.net/Charts/PDR.htm#FujiFilm%20X-T5,Sony%20ILCE-7RM4
That's good news. As is readable in Bill's graphs,
For my purposes both are overkills IQ wise. Sometimes ISO50 and PL-filter is better/easier than ND filters.
PDR has no reference to any exposure or "middle grey" that is at a nominal exposure for the ISO setting.
What those trends say to me is:
Both cameras get their three lowest ISO settings by using the same gain/digitization as their 4th, 5th and 6th ISO settings, but with 1 stop less headroom, by using higher raw levels for greypoints.
A higher conversion gain occurs on the A7R4, starting at ISO 320.
Both cameras end analog gain at ISO 12800.
Above ISO 12800, the Fuji uses the ISO 12800 gain and digitization, and just moves the greypoint and whitepoint to lower raw values for the two higher ISOs, while the Sony uses multiplication of raw values, tossing extended digitized headroom into the trash more at the highest ISOs. The Sony has higher PDR at ISO 16000 and up than expected from the trend below 16000, but this is due to raw cooking.
Thank you. I never did think that ISO 50 has effect in jpeg only. But so it is.
I took test shots in manual mode with same exposure settings in ISO 50 and 100. The raw histograms
in RawTherapee were identical. I even could copy/paste my developing settings from ISO 50 shot to
ISO 100 picture to get same tones and rgb histogram in tiff.
My user's manual explains the use of extended ISO settings, but says nothing about them being "special." I don't use them, and the camera hasn't exploded yet.
