Sorry, I didn't read carefully enough. On a side note, if my memory isn't failing me, the SD9 actually doesn't change pre-ADC gain with ISO, so is 'ISOless'.
You can control the ratio of ISO setting to exposure in both full manual and auto-ISO M with "EC" (ISO bias). The differences are that "EC" usually has a more limited range, often just +/- 3 stops, whereas full manual can allow a lot more (up to the limits of the ISO range), and full manual is static while auto with "EC" responds to metering. So, when lighting is static, full manual may be more useful, but when lighting changes rapidly, the dynamic response to metering may be much safer.
Being someone who shoots subjects that jump from sunlight to shade and back to sunlight again within a couple of seconds, where it takes 120% concentration just to keep the subject in the frame, full manual seems a bit ridiculous for such situations, unless you don't mind the extra read noise and just let the shade shots be very dark OOC and potentially have more read noise (that's not something that varies with your SD9, but it does with most cameras). Anyway, the potential error/damage is greatest in such situations when full manual is employed and the settings are optimized for the shade. If you choose a higher ISO for a brighter OOC and review image and less read noise in the shade, as soon as you point your lens back onto something sunlit, you are all clipped to white, whereas, auto-ISO would have dropped the ISO setting to save you from the harsh clipping.
That certainly makes some kinds of decision-making simpler (all you suffer is a dark default conversion when the ISO is much lower than the ISO exposure index), but it is not an absolute quality. When digital photo culture started talking about "ISO-invariance" and "ISO-less-ness", it seems that there was a lot of assumption that this was the Holy Grail in the crusade against read noise, but these qualities are self-relative within a camera, and have nothing to do with absolute input-referred read noise. Historically, the most ISO-invariant cameras are often the ones with the most pre-gain read noise. Sigma SD9 and Canon G9 are two good examples. The SD9 uses no variable gain, and the G9 uses it but it has no effect on visible noise, and they are both very noisy cameras. You can take just about any current camera with similar sensor size and "under-expose" base ISO to simulate the SD9 or G9, and get a lot less noise at all ISO exposure indices.
No problem! Yes, it is ISO-less. Along with most but not all pre-'Quattro' models, the ISO setting is passed in meta-data to the raw converter on the computer.
I still don't understand "all you suffer is a dark default conversion when the ISO is much lower than the ISO exposure index" as regards my Sigma SD9 ...
Most of the CCD cameras were ISOless. They did use pre-ADC variable gain, but the read noise was high enough to mask ADC noise with a 12-bit ADC, so the variable gain didn't help. In those days it was more true (though still wrong) to say ISO was gain, because the gain was being used as an analog computing multiplier, multiplication being an expensive operation using the processors they had at the time, which were slow, 16 bit only and didn't have single cycle multipliers. As a long-time Canon user, it probably didn't affect you.
I find unless you are shooting a once in a lifetime scene , where I will use every trick in the book to milk every drop out of the camera . That manual with auto ISO does a surprisingly good job set the aperture you need for DOF and shutter speed for subject movement and let auto ISO take care of things . I think Nikon's implementation of auto ISO works well
Oooh, scary, I'm so used to thinking of DOF in f8 . 😆
I'l have to try it, though. I've become a fan of Nikon algorithms in white balance and highlight weighting, why not add another abstraction to my point-n-shooting...
Yes! I personally am on the side of maximum DOF most of the time, which is one reason why I prefer short FL lenses at small aperture, and am happy that today one can avoid most of any concommitant noise due to high ISO settings.
I wonder what the ratio of “f/64 lovers” to “bokeh lovers” is.
The problem with 'maximum DOF' is that even if you like the effect, it degrades image sharpness in addition to the noise effects. You see many photos where the aperture is much smaller than is needed to provide DOF for the whole of the visible scene.
It's not really a case of that simple a dichotomy. Different shots call for different treatments - I think that a little bit of differential focus is a very useful compositional tool - it's not always that one needs to go to either extremity.
I'm think it always depends on what you're shooting. In the high magnification macro sphere f/32 still means shallow DOF but with significantly worse image quality, so it's not a great option most of the time. For landscapes or architectural shots more DOF might be beneficial and the downsides often not relevant or even visible.
I personally prefer "shallow" (or at least visibly limited) DOF in a vast amound of cases. I even love tilting/rotation the plane of focus when I'm shooting landscapes, because it helps me focus on certain things, but I know that I'm in the minority there:
So I'm certainly in the group of "bokeh lovers" even though it's very important to me that there's enough detail in what I want to draw attention to. I get your different preference though and will try to keep it in mind when I'll have more time outside shooting at distance!
Simplejoy: I agree that the preference depends upon the subject matter, and my usual subject matter, which is not the same as yours, fits the "max in focus" method. That said, I do not have the patience to stack images!
I like your castle photo above, though I pine for being able to see what lies behind!
I use Auto-ISO with Manual Exposure mode most of the time, and straight manual for ISO 64 or 100 shots more often than Aperture Exposure mode (again, usually in conjunction with Auto-ISO). On my Nikon DSLRs Auto-ISO is a straight up metering issue, and the Nikon DSLRs I've used since the D200 by default tend to have a good deal of headroom in the default settings. That being the case, I will routinely set EC between +.3 to +1 using Center-weighted metering (set to "Average") as my quick and dirty metering of the scene, but when using Auto-ISO I set EC back to zero or even a negative value if the scene has a lot of brighter areas that may get "averaged" out of the sensor's latitude (i.e., they will end up overexposed).
With my D850 I might just set ISO to 64 and forget about changing the ISO setting, but that would make chimping and showing the first takes on SnapBridge impractical if not impossible. As I see it, raising ISO does no harm as long as highlights are preserved.
Thank you.
BTW, for me "full frame" simply means I'm not cropping the shot.