JD750 Loc: SoCal

Do you mean my camera will be offended more easily at lower ISO? Oh no! I don't want that.

selmslie Loc: Fernandina Beach, FL, USA

According to the information at Exposure value you need an effective EV of between 8 and 9. You can achieve that at f/5.6 and 1/500 sec with an ISO of between 3200 and 6400.

If that's too noisy you will need to lower the shutter speed and/or open the aperture in order get more exposure and use a lower ISO.

If you want to use ISO 200 you may have to go as low as 1/125 sec at f/2. That may be too slow and not enough DoF.

blackest Loc: Ireland

Doesn't it really come down to there is a cost to not using enough light?

In camera, the signal can be boosted to get the values needed to produce a "good" jpeg. If you are shooting raw you don't actually need that jpeg. You can do your own boosting in Post.

Isn't the camera really saying when it suggests say 1600 iso is thats the iso factor I need to apply for your mid grey tone to be exposed correctly. It really has no interest in the shadow regions.

Maybe its better to use an incident light meter or maybe a spot meter to figure out the exposure needed to keep the signal above the noise in the dark areas. If you can do that then it may be under exposed but you can correct it cleanly.

As for iso invariance if you look at the dxo charts plotting dr vs iso if they are a straight line then thats going to be iso invariant if its more like a staircase then its likely there is an adjustment in analog gain of the signal. If your iso is falling in that staircase region you probably will lose iq if you under expose.

Seems the advantage of under exposing is that you ensure no highlight detail is blown but there can be a cost if the shadow detail isn't recorded above the noise floor.

CatMarley Loc: North Carolina

It will always be a compromise. And the question always be: "which set of compromise settings will give me the best result?" The answer to this question, for the present at least, will depend on the subject, the available or artificial light and the attributes of the particular camera and lens. In the future I suspect we can expect more sensitive sensors, with better s/n and DR and the set of compromises will all change. Meanwhile it is worth some experimentation to find out the limits of what works with your particular equipment.

selmslie Loc: Fernandina Beach, FL, USA

If you have the right exposure to produce a good JPEG straight out of the camera you are probably going to have plenty of information in the raw file as well.

But if the JPEG is too bright or dark than you are going to have to do some work to get a good final image. If it's even a little too bright you might have blown highlights but there is more latitude on the dark side.

Toby

Gee common sense suggestion ???????????

Notorious T.O.D. Loc: Harrisburg, North Carolina

Some good thoughts but the problem with suggesting an incident light meter is that most people just don't have one nor want to bother with one. So many people are quite happy with the phone photo they can take anytime and place... Some of us are seeking more...

Notorious T.O.D. Loc: Harrisburg, North Carolina

And lets not forget that it is also hopefully the result that the shooter is aiming to achieve. Mostly I am shooting to document the subject at a time and place but sometimes I a looking to do something very different...

a6k Loc: Detroit & Sanibel

Isn't it the opposite? Isn't Tri-X more sensitive (ASA 400) than Plus-X (ASA 125)?

selmslie Loc: Fernandina Beach, FL, USA

That's an apples and oranges comparison.

Your sensor's sensitivity does not change when you change the ISO setting. What changes is the amplification or gain applied to the charge recorded in the sensor.

You can expose film at any ISO/ASA you chose - at your own risk. You can also use minus development on Tri-X or plus development on Plus-X Pan (discontinued) or FP4+ and shoot them both at ISO 200. This will affect the contrast in either negative and the density of the highlights. What you can't do much about is how much gets recorded at the low end of the film curve.

You can read more on this at Film vs. Digital Characteristic Curves.

CatMarley Loc: North Carolina

Here's an accidental underexposure, pushed 2EV. Glad I was able to save this one of my sweet little girl.

(Download)

a6k Loc: Detroit & Sanibel

Scotty, I see your point about apples and oranges but the real question remains. Isn't a sensor with a "native or base" ISO of X less sensitive than one with ISO of 2X? I understand that the camera can be set to amplify the signal by increasing the ISO setting but the base ISO represents the inherent sensitivity (or does it?).

Thanks for your patience with my clumsy and misleading analogy.

DaveC1 Loc: South East US

This is a really good article.

JD750 Loc: SoCal

There is a lot more to processing an image than the the sensor. It starts with the sensor but there is way more happening after that. The quality of the amplifier and the processing steps are all important, and there is processing involved even on a raw file. A good analogy is audio. Way more goes into to a recorded song than the microphone. The microphone is important but it's not the only thing that matters.

selmslie Loc: Fernandina Beach, FL, USA

The sensor itself has exactly the same sensitivity regardless of the ISO.

For example, you might fill up the buckets (sensels) to 1/4 of their capacity at ISO 400 with a given exposure (1/60 @ f/11, EV 13) and it might look 1 stop overexposed (on the edge of having blown out highlights). The same exposure at ISO 200 might look correctly exposed and it might be 1 stop below being blown out. At ISO 100 it might look underexposed and two stops below blowing out.

The corresponding values recorded in a 14-bit raw file would be about 16000, 8000 and 4000.

My earlier statement about a sensor that can use a lower ISO was not quite what I meant. What I meant to convey was that it has a greater capacity - deeper "buckets".