I've not read all the posts of this very interesting thread involving great and informed contributers.
From a number of posts, am I wrong when I find that the controversy materialize in
practice in the following question :
ETTR or High ISO ? provided :
- High ISO (with analogue gain) lowers information quantization issues by augmenting values earlier in the overall processing pipeline, in an analogue way (continuous rather than discrete way). But..
- High ISO does reduce Dynamic Range, as charts show
It's the limited volume of light that reduces dynamic range; not the ISO.
In the below chart comparing the dynamic range of the iPhone 7 and X-T3, the Fuji's dynamic range matches the smartphone's only when the volume of light delivered to the APS-C sensor matches that of the much smaller iPhone sensor at full saturation.
IMVHO, many questions here resume in that ones :
- Does ETTR make sense at a higher than base ISO ?
- If there is not enough light to "saturate" the sensor given minimal SS and maximal aperture of the lens, then what's the point of ETTR ?
Keeping in mind the whole purpose of ETTR is to maximize the volume of light delivered to the sensor, I would argue that within the limitations of the scene brightness, f-stop and shutter speed needed to realize the photographer's creative vision, it always makes sense to maximize the light delivered to the sensor.
The more "information/data" you have to work with, the better the resulting image. This approach does not always correspond to a histogram display indicating a majority of pixels ETR. But again, the purpose of ETTR is not to achieve a certain histogram profile; it is to maximize the light delivered to the sensor.
Understanding the answers, just adapt your policy depending of the shot.
I basically do as Doug writes :
"you select the lowest ISO that gives you acceptable exposure settings (shutter speed and aperture)" (just taking care of the 800 ISO thing duer to bi-amp) and I only need 3 ISO settings :
- if light is abundant enough provided my desired SS and aperture, then if possible I use base ISO 200 and ETTR (usually giving overecposed JPEGS but correctly exposed RAW files, technically speaking of course)
- if light is not so abundant, then I use Auto ISO. I have 2 Auto ISO settings (one from ISO 200 to 640 and one over 800)
I'm not sure you do as Doug writes or even that Doug does as he wrote. The first bullet point, above, suggests you have a desired f-stop and shutter speed in mind, and that you use ISO as an indicator of the volume of light delivered to the sensor. My guess is that Doug does much the same. It's a sensible approach given that exposure settings tend to have more direct impact on one's ability to realize a creative vision for a photograph.
However, the statement quoted and italicized, above, is written to give priority to choosing an ISO, first, and then selecting f-stop and shutter speed to optimize exposure. I suspect that was not the impression Doug intended to give.
In most situations, it's an approach that makes little sense. ISO, in and of itself, is not an end. We don't select ISO as a
desired outcome; at least, not in the same way we select a desired depth of field (f-stop) or a desired extent to which motion is frozen (shutter speed). These are outcomes that contribute to realizing an envisioned image. ISO is not.
--
Bill Ferris Photography
Flagstaff, AZ