at the lowest possible ISO (in regard to read noise (...)) to get the largest possible DOF to minimise highlight clipping. Which essentially means you expose for highlight protection and to balance noise against blur (low DOF, shake, motion). So, yes you expose with considerations about bright areas and general noise, but you do NOT expose for a mid grey.
You avoid the point by rambling on with something which makes little sense. DOF does not minimise highlight clipping.
That was a typo, I meant DR. The lower the ISO, the higher the DR. But since a significant portion of cameras have increasing read noise with decreasing ISO, you have to balance DR (or highlight protection) against noise in the darker areas. Essentially you can have with these cameras for a given exposure more highlight protection (lower ISO setting) or lower dark area noise (higher ISO setting).
But that is a finer point. In a sense, it is the exception from the rule that exposure is 'only' about balancing maximising exposure vs. highlight clipping. But trying to explain you the exception before you have internalised the rule is highly unlikely to work, so maybe better disregard this aspect for the time being.
What are "considerations about bright areas"? Sounds to me like you're trying to avoid saying that you meter. Or are you going to claim you don't use metering?
Now, what could I have meant with "considerations about bright area"? Since we have been talking about highlight clipping, maybe I was just rephrasing things...
And BTW, I expose to set mid grey where I want it and so do you. This does not preclude me from taking meter readings from the highlights or shadows.
You know full well that if you blow a highlight, it is gone. But if your mid grey does not land exactly at {128} it is a piece of cake to move it there without any real impact on highlight clipping. Why would you thus care more about placing the mid grey during exposure than selecting exactly which part of scene is clipped? You cannot place mid grey exactly where you want it and clip exactly a chosen amount of highlights by optimising exposure, moving the mid grey moves your clipping point. Thus shouldn't your first consideration when choosing exposure be what highlight get clipped and which don't? And if noise is more (or equally) important than highlight clipping, shouldn't your first consideration be about maximising exposure (and thus minimising photon shot noise)?
Why would would you prioritise something you can change in post over something you cannot change in post (highlight clipping & noise)?
So let's analyse what you do. You choose aperture, shutter speed and ISO, you don't say how you come to this particular combination ( 'rough' considerations about bright areas perhaps? - doesn't the D3's metering system work? ).
The metering system gives me a rough first shot of what exposure I set (though I know roughly what the meter will tell me in common light scenarios), the histogram is the next step. It tells me about highlight clipping (not absolutely perfectly but UniWB helps with that).
a) mid grey is where you want it to be and therefore you have a 'correct' exposure and you will not need to correct it in PP, or
b) mid grey is not where you want it to be and therefore you have an 'incorrect exposure' which you will correct in PP (assuming, of course, that you have avoided clipping wanted highlights or blocking wanted shadow detail).
I tend to white-balance and 'exposure'-correct every single photo I 'publish' (ie, show to others), though I try 'copy' WB settings to sets of pictures if I know that the lighting was constant, though I still am constantly tempted to fine-tune it. 'Exposure'-correction usually means 'exposure', brightness and contrast slider, though often one or two of them will do.
So, whenever you use the word 'exposure' you actually mean exposure + image processing? Is that in your estimation what photographers generally mean when they say exposure?
No I don't, I mean the amount of light captured for a given ISO speed - the exposure triangle.
Do you really believe that changing the ISO setting changes the amount of light being captured? Light is captured by the photodiodes in every pixel, the signal is then read out, optionally amplified before being converted into a digital signal. Changing the ISO setting does in no way affect how the light creates charge in the photodiodes.
Moreover, since changing the ISO setting is just another fancy way of instructing the camera (and indirectly the raw processor) how to process the data read off the sensor, you actually
do mean exposure + image processing.
You just have a hard time (like a lot of people), to accept that changing the ISO setting is mainly just about changing how the sensor data are processed into an image. The ISO setting has nothing to do with exposure (except that changing it will cause your camera to
suggest you a different exposure) unless you take the creative position that exposure is everything that affects what data comes out of the A/D converter.
It's not a question of learning but what you learn...
So, what new have you learned about digital sensors in the last couple of years?
