Exposure is not how dark or how light an image looks!

Under- and over-exposure are terms by themselves; they do not contradict or change the meaning of exposure. As Jim and a few other mentioned, they mean being below or above the intended exposure.

It depends on the meaning of exposure someone uses. Many regard exposure as just how light or dark an image looks and so they see over/under exposure as meaning if in their opinion the image looks too light or too dark regardless of the exposure* that was set.

I could have set the optimal exposure** but the image could still be too dark if the ISO was set too low. In this case the image is not under exposed*** because I could not have added more exposure* without messing up my DOF or blur requirements. The image is just simply too dark.

In any case, for the sake of clarity I include the definitions of 'exposure', 'optimal exposure' and 'under exposed' in my signature so everyone can at least see what I mean when I use the terms in posts.

* exposure - amount of light striking the sensor per unit area while the shutter is open
** optimal exposure - the maximum exposure* within dof and motion blur requirements without clipping important highlights.
*** under exposed - more exposure* could have been added with the DOF and blur constraints still being met without clipping important highlights.

And not just in how the print looks but in how the image looks on someone else's screen as well.

I sense that many people incorrectly assume that how an image looks on there screen is exactly how it looks on the image creator's screen which in many cases could be significantly different.

For example, if someone has their screen brightness set way higher than mine is then obviously my images, although looking perfect on my screen will most probably be too light on the screen of the person viewing my image. It is technically incorrect for that person to say my image is over exposed unless they were certain I used a greater than optimal exposure**.

All the other person can say with accuracy is that my image looks too light on their screen for their taste.

I don't understand the point you're trying to make.

thank you , the article constantly references "recording medium" with digital we have a "variable recording medium" which now needs to be set as the 3rd adjustment for exposure to be classified as correct exposure/image brightness for a given scene representing the photographers intent and purpose of the image. ;-) that was a mouth full :-)

Your quote " It is common for photographers to omit scene luminance when talking about exposure.

Q: "What exposure did you use for that shot?"

A: "I gave it 1/125 at f/8."

its not common for any photographer to omit iso or lighting given ambient or artificial. no one ever talks just exposure, because it has no meaning by itself.

It's not an article, just a forum post. For more depth, I've previously linked an excellent and informative article; but that's still just a beginning.

You'll find "recording medium" three times, once in each paragraph. You need a medium to record the light values — or the exercise is all in vain. Either film or an electronic image sensor will work for this discussion.

We have variable ISO ratings and settings for film and image sensors...

...which have always been part of the calculation to determine exposure value.

That's always been the end goal.

Photographers may include or omit many things depending on the current topic of discussion and context.

Sure it does. "1/125 at f/8" represents an exposure value of EV 13. That's the approximate exposure value for ISO 100 on a cloudy day.

Light meters don't need to be perfect, just reasonably accurate and consistent. They generally work very well and, as the user's understanding of them improves over time, any good meter can produce reliable results.

Why would that be a problem? Each one may have had different intentions but so what?

Indeed. I use my camera's highlight-weighted matrix metering most of the time, which automatically kaboshes any (currently-available) automatic use of a tone curve. I've accommodated this in my post workflow, so it's not a big deal.

I think it is quite common for photographers to omit scene luminance when discussing exposure. ISO is properly omitted, because ISO has no effect on exposure. Lastly, exposure by itself has great meaning, if you understand how it works.

I think it is sad that so many photographers shooting raw and trying to get the best exposures don't ever look at their files with a program like RawDigger. If you don't know what the raw data looks like, you are just winging the exposures. I don't think you have to look at the raw data for every image, but I think you should routinely look at it so you can see how well you're doing.

Edit: not sure why this was posted twice. The first time I got an "unknown error" message.

It depends on your goal. If your concern is the underlying numbers, then RawDigger is a useful tool. If your goal is producing an image with no visible noise, then your eyes may be able to tell you whether or not you have reached your goal.

In some sense it’s like putting water in a bucket. Once you have put enough water into the bucket, pouring more in doesn’t make the bucket more full. Once in image reaches the point where there is no visible noise, further reductions in noise don’t change the appearance.

If you have a five gallon bucket, pouring 6 gallons of water gives you the same result as pouring in 60 gallons. You end up with a bucket that is full. It’s not that 60 gallons gives you a better result. Once your image quality is so good that there are no visible issues, then you can make a reasonable argument that further measurable improvements yield the same result, not a better result.

By (my) definition, it is the exposure that met their intentions.

So if I say that under or over exposure is the one which was too low or too high to meet their intentions, you would be fine with that?

You could be routinely underexposing from ETTR by two stops and getting noise quality that you consider acceptable. If that were true, you could get the same noise with a smaller sensor and save your self a lot of money.

You could see noise in your image and be unsure about how to fix it. Looking at the raw data could tell you why your image is noisy and what to do about it.

I can't count the number of times people have told me that they are masters of ETTR and they know it by looking at the histogram in Lr.