Exposure with digital cameras

If the entire image is clipped highlights (or shadows), there is no data to recover, hence the white image just turning gray. That's not really a fair example.

Here are some images showing an almost 5-stop processing range. There is a small amount of clipping on each end, but not enough to make a visual impact (so it is not quite a full 5 stops).

The first panel shows three images: -3 EV, 0 EV, and +2EV. These are RAW files processed in LR with "daylight" WB and no other adjustments. (All images: Nikon D850, Sigma 20mm f/1.4 taken within 70 seconds of each other at f/16, ISO 64, and 1/100, 1/13, and 1/3 seconds.)

View attachment cf6132d5b78e4fe6998b51c4aaa6586d.jpg
Exposure compensation: -3 EV, none, and +2 EV (right).

The next panel shows the same images with the exposure corrected (+3 EV, 0, -2 EV):

View attachment ee0c0ee177f94b25b5c4c2f0b71dfdac.jpg
Each image adjusted in LR: +3 EV, 0, and -2 EV (right).

Looking at them in full size, there are some clear (but minor) differences in color, but they are pretty similar overall.

I agree with your main point: watch for clipped highlights--no recovery. But there is a lot of data hiding in the shadows!
 
If the entire image is clipped highlights (or shadows), there is no data to recover, hence the white image just turning gray. That's not really a fair example.

Here are some images showing an almost 5-stop processing range. There is a small amount of clipping on each end, but not enough to make a visual impact (so it is not quite a full 5 stops).

The first panel shows three images: -3 EV, 0 EV, and +2EV. These are RAW files processed in LR with "daylight" WB and no other adjustments. (All images: Nikon D850, Sigma 20mm f/1.4 taken within 70 seconds of each other at f/16, ISO 64, and 1/100, 1/13, and 1/3 seconds.)

View attachment cf6132d5b78e4fe6998b51c4aaa6586d.jpg
Exposure compensation: -3 EV, none, and +2 EV (right).

The next panel shows the same images with the exposure corrected (+3 EV, 0, -2 EV):

View attachment ee0c0ee177f94b25b5c4c2f0b71dfdac.jpg
Each image adjusted in LR: +3 EV, 0, and -2 EV (right).

Looking at them in full size, there are some clear (but minor) differences in color, but they are pretty similar overall.

I agree with your main point: watch for clipped highlights--no recovery. But there is a lot of data hiding in the shadows!
But undoubtedly the -3 (then +3 correction), has more "noise" if that needs to be a significant consideration. (and the ETTR +2 to -2 the least noise)

A very good test ... Thank You ... and a good (average) subject since a nice range of "white" clouds to (shadowed/dark) trees in image.

BUT HEY ... can you do this test again and indeed include +3 & +4 (& +5) in the original series ???

I suspect that at some point the "whites", (clouds) would indeed be un-recoverable.
 
Last edited:
The last couple of days in a discussion in the beginners forum, it had repeatedly been stated that getting an exposure with a digital camera isn't as critical as it was with shooting film.

One such quote ...

"With film, a "good exposure" was one that fell into the narrow range needed to produce a usable negative. If the exposure was too high, or too low, the negative wasn't usable (too thin, or too dense). With digital, the range of exposures that give a very usable result is well over 5 stops"

Today I decided to test this out. First I took in exposure with 0 EV on the dial...
You misunderstand the word "exposure". That's the amount of light that falls on the sensor. ISO is not a factor in exposure, however it is a factor in image lightness.

Try repeating the test, but this time don't pretend you are using a film camera - accept that you are now shooting with a digital camera. Your camera doesn't have any film in it, don't use a fixed ISO!

Try the following test:

Use a meter to determine a reasonable f/stop and shutter speed for ISO 100.

Put your digital camera in manual mode and select the meter recommended shutter speed and f/stop.

Remembering that this is not a film camera, don't select a fixed ISO. Set the ISO to AUTO!

Now take a test shot. Now reduce the "exposure" (light reaching the sensor) by one stop (faster shutter speed or smaller aperture). Take another photo.

Reduce by another stop and repeat.

Take 5 or 6 test shots, each with one stop less light than the previous shot.

What you should find is that your digital camera gives you quite usable images with good image lightness across that 5 stop exposure range.

====

Remember, the ISO setting affects the mapping from exposure to image lightness. Digital cameras have the advantage that the camera can vary this from frame to frame. With film, you were stuck with the same ASA for the entire role.

Of course, if you manually set an ISO that doesn't match your exposure, your out of camera JPEG won't look good. But why would you want to limit your digital camera to the abilities of a film camera?
 
The last couple of days in a discussion in the beginners forum, it had repeatedly been stated that getting an exposure with a digital camera isn't as critical as it was with shooting film.

One such quote ...

"With film, a "good exposure" was one that fell into the narrow range needed to produce a usable negative. If the exposure was too high, or too low, the negative wasn't usable (too thin, or too dense). With digital, the range of exposures that give a very usable result is well over 5 stops"

Today I decided to test this out. First I took in exposure with 0 EV on the dial...
You misunderstand the word "exposure". That's the amount of light that falls on the sensor. ISO is not a factor in exposure, however it is a factor in image lightness.
I am curious about your understanding since you keep coming back to comment that others misunderstand the word "exposure".

When you adjust your digital camera to make an Exposure Compensation are you changing the exposure? If not, what are you changing, aperture value, shutter speed, or ISO or all of the above?
Try repeating the test, but this time don't pretend you are using a film camera - accept that you are now shooting with a digital camera. Your camera doesn't have any film in it, don't use a fixed ISO!

Try the following test:

Use a meter to determine a reasonable f/stop and shutter speed for ISO 100.

Put your digital camera in manual mode and select the meter recommended shutter speed and f/stop.

Remembering that this is not a film camera, don't select a fixed ISO. Set the ISO to AUTO!

Now take a test shot. Now reduce the "exposure" (light reaching the sensor) by one stop (faster shutter speed or smaller aperture). Take another photo.

Reduce by another stop and repeat.

Take 5 or 6 test shots, each with one stop less light than the previous shot.

What you should find is that your digital camera gives you quite usable images with good image lightness across that 5 stop exposure range.

====

Remember, the ISO setting affects the mapping from exposure to image lightness. Digital cameras have the advantage that the camera can vary this from frame to frame. With film, you were stuck with the same ASA for the entire role.

Of course, if you manually set an ISO that doesn't match your exposure, your out of camera JPEG won't look good. But why would you want to limit your digital camera to the abilities of a film camera?
 
The last couple of days in a discussion in the beginners forum, it had repeatedly been stated that getting an exposure with a digital camera isn't as critical as it was with shooting film.

One such quote ...

"With film, a "good exposure" was one that fell into the narrow range needed to produce a usable negative. If the exposure was too high, or too low, the negative wasn't usable (too thin, or too dense). With digital, the range of exposures that give a very usable result is well over 5 stops"

Today I decided to test this out. First I took in exposure with 0 EV on the dial...
You misunderstand the word "exposure". That's the amount of light that falls on the sensor. ISO is not a factor in exposure, however it is a factor in image lightness.
I am curious about your understanding since you keep coming back to comment that others misunderstand the word "exposure".

When you adjust your digital camera to make an Exposure Compensation are you changing the exposure? If not, what are you changing, aperture value, shutter speed, or ISO or all of the above?
Exposure-Compensation can be referred to as a "BRIGHTNESS" adjustment.

It changes/offsets the "metering" calibration, (which does indeed also change the exposure-settings).

A common use is to change the (normal) gray of a SNOW scene to a (correct) "white".

And change the (normal) gray of a BLACK (night) sky to a (correct) "black".

It can also be used to lighten shadowed/back-lit silhouettes -- and -- darken (too-bright) spot-lit subjects on stage, (w/ darker background).

Note that EC can be the most-important and most-used control on your camera, (but often mis-understood).
 
Last edited:
The last couple of days in a discussion in the beginners forum, it had repeatedly been stated that getting an exposure with a digital camera isn't as critical as it was with shooting film.

One such quote ...

"With film, a "good exposure" was one that fell into the narrow range needed to produce a usable negative. If the exposure was too high, or too low, the negative wasn't usable (too thin, or too dense). With digital, the range of exposures that give a very usable result is well over 5 stops"

Today I decided to test this out. First I took in exposure with 0 EV on the dial...
You misunderstand the word "exposure". That's the amount of light that falls on the sensor. ISO is not a factor in exposure, however it is a factor in image lightness.
I am curious about your understanding since you keep coming back to comment that others misunderstand the word "exposure".

When you adjust your digital camera to make an Exposure Compensation are you changing the exposure? If not, what are you changing, aperture value, shutter speed, or ISO or all of the above?
Exposure-Compensation can be referred to as a "BRIGHTNESS" adjustment.

It changes/offsets the "metering" calibration, (which does indeed also change the exposure-settings).
So it does change exposure settings among other things? Exactly what?
A common use is to change the (normal) gray of a SNOW scene to a (correct) "white".

And change the (normal) gray of a BLACK (night) sky to a (correct) "black".

It can also be used to lighten shadowed/back-lit silhouettes -- and -- darken (too-bright) spot-lit subjects on stage, (w/ darker background).

Note that EC can be the most-important and most-used control on your camera, (but often mis-understood).
 
The last couple of days in a discussion in the beginners forum, it had repeatedly been stated that getting an exposure with a digital camera isn't as critical as it was with shooting film.

One such quote ...

"With film, a "good exposure" was one that fell into the narrow range needed to produce a usable negative. If the exposure was too high, or too low, the negative wasn't usable (too thin, or too dense). With digital, the range of exposures that give a very usable result is well over 5 stops"

Today I decided to test this out. First I took in exposure with 0 EV on the dial...
You misunderstand the word "exposure". That's the amount of light that falls on the sensor. ISO is not a factor in exposure, however it is a factor in image lightness.
Well as a photographic exposure is x amount of light per unit area derived from the scene luminance at a certain f/ratio for a given exposure time... f/6.4 for 1/640 sec vs f/6.4 for 2 sec of the same scene, are different exposures.
Try repeating the test, but this time don't pretend you are using a film camera - accept that you are now shooting with a digital camera. Your camera doesn't have any film in it, don't use a fixed ISO!

Try the following test:

Use a meter to determine a reasonable f/stop and shutter speed for ISO 100.

Put your digital camera in manual mode and select the meter recommended shutter speed and f/stop.

Remembering that this is not a film camera, don't select a fixed ISO. Set the ISO to AUTO!

Now take a test shot. Now reduce the "exposure" (light reaching the sensor) by one stop (faster shutter speed or smaller aperture). Take another photo.

Reduce by another stop and repeat.

Take 5 or 6 test shots, each with one stop less light than the previous shot.

What you should find is that your digital camera gives you quite usable images with good image lightness across that 5 stop exposure range.
So what you're actually saying is, cameras these days have a usable ISO range of 5 stops. Someone could use anything from ISO 100 to 3200 and achieve an acceptable photo.
 
When you adjust your digital camera to make an Exposure Compensation are you changing the exposure?
It is easier to think that exposure compensation alters the way your camera measures the brightness of the scene, and estimates the correct exposure. Used to be that it simply averaged all light, and presumed it was supposed to be gray -- hence, if you were photographing a snowy landscape, you would have to adjust the exposure compensation. In effect to tell your camera, do not assume the scene is gray, I am telling you it is white.

In modern cameras the exposure calculation is more sophisticated, but they do still get it wrong. If your whites turn gray in your photos, you need to increase the exposure compensation.
If not, what are you changing, aperture value, shutter speed, or ISO or all of the above?
Exposure compensation only changes things when in automatic modes. In manual, it does nothing. What is being changed depends on the automatic mode you are using. If you are in aperture priority mode, then shutter speed is being changed. If you are in shutter priority, aperture is being changed. If and how ISO is being changed depends on your auto ISO settings.
 
Last edited:
The last couple of days in a discussion in the beginners forum, it had repeatedly been stated that getting an exposure with a digital camera isn't as critical as it was with shooting film.

One such quote ...

"With film, a "good exposure" was one that fell into the narrow range needed to produce a usable negative. If the exposure was too high, or too low, the negative wasn't usable (too thin, or too dense). With digital, the range of exposures that give a very usable result is well over 5 stops"

Today I decided to test this out. First I took in exposure with 0 EV on the dial...
You misunderstand the word "exposure". That's the amount of light that falls on the sensor. ISO is not a factor in exposure, however it is a factor in image lightness.
I am curious about your understanding since you keep coming back to comment that others misunderstand the word "exposure".

When you adjust your digital camera to make an Exposure Compensation are you changing the exposure? If not, what are you changing, aperture value, shutter speed, or ISO or all of the above?
Exposure-Compensation can be referred to as a "BRIGHTNESS" adjustment.

It changes/offsets the "metering" calibration, (which does indeed also change the exposure-settings).
So it does change exposure settings among other things? Exactly what?
Digital cameras are not film cameras. However, manufacturers tried to make them mimic film cameras, because that's what photographers were used to.

On a digital camera the setting called "Exposure-Compensation" is actually "Lightness-Compensation". It adjusts the camera's metering system to change the target lightness for the final image.

If your camera is in manual mode with a fixed ISO, then Exposure-Compensation is decoupled from both exposure and the mapping from photons captured to image lightness. In this mode changing Exposure-Compensation doesn't directly affect the image, it merely affects the feedback the metering system is giving you.

If the camera is in an automatic mode, the camera will adjust what you have allowed it to adjust in order to affect image lightness. This might be a change to aperture, shutter speed or ISO.

In Auto-ISO mode, Exposure-Compensation may result in only a change to ISO, and not an actual change to exposure.

====

One of the issues here is that there are multiple settings which affect image lightness.
  • f/8, 1/125, ISO 200
  • f/4, 1/250, ISO 200
  • f/8, 1/250, ISO 400
All yield the same image lightness, yet the last one has a different exposure.

Digital is not film. There is no need to treat it as such. There is not need to pick an ISO before the shoot, and keep it fixed for the every shot on that SD card. Not only can you vary ISO whenever you like, you can allow the camera to select ISO in the same way that you can allow it to select shutter speed and aperture. They are all components of image lightness.
 
The last couple of days in a discussion in the beginners forum, it had repeatedly been stated that getting an exposure with a digital camera isn't as critical as it was with shooting film.

One such quote ...

"With film, a "good exposure" was one that fell into the narrow range needed to produce a usable negative. If the exposure was too high, or too low, the negative wasn't usable (too thin, or too dense). With digital, the range of exposures that give a very usable result is well over 5 stops"

Today I decided to test this out. First I took in exposure with 0 EV on the dial...
You misunderstand the word "exposure". That's the amount of light that falls on the sensor. ISO is not a factor in exposure, however it is a factor in image lightness.
Well as a photographic exposure is x amount of light per unit area derived from the scene luminance at a certain f/ratio for a given exposure time... f/6.4 for 1/640 sec vs f/6.4 for 2 sec of the same scene, are different exposures.
Try repeating the test, but this time don't pretend you are using a film camera - accept that you are now shooting with a digital camera. Your camera doesn't have any film in it, don't use a fixed ISO!

Try the following test:

Use a meter to determine a reasonable f/stop and shutter speed for ISO 100.

Put your digital camera in manual mode and select the meter recommended shutter speed and f/stop.

Remembering that this is not a film camera, don't select a fixed ISO. Set the ISO to AUTO!

Now take a test shot. Now reduce the "exposure" (light reaching the sensor) by one stop (faster shutter speed or smaller aperture). Take another photo.

Reduce by another stop and repeat.

Take 5 or 6 test shots, each with one stop less light than the previous shot.

What you should find is that your digital camera gives you quite usable images with good image lightness across that 5 stop exposure range.
So what you're actually saying is, cameras these days have a usable ISO range of 5 stops. Someone could use anything from ISO 100 to 3200 and achieve an acceptable photo.
"Acceptable" in terms of image lightness. I was not commenting on other aspects of the image.

For instance, cameras can take images over a wide range of shutter speeds and apertures. The images may have acceptable lightness, but may be unacceptable due to other issues (the depth of Depth of Field may not be sufficient to get your entire subject in focus, diffraction issues may cause an unacceptable lack of sharpness, motion blur might be too high, etc.)

As photographers we make choices which affect the final image. Most of these choices affect multiple aspects of the image. For instance aperture affects exposure and depth of field. Shutter speed affects motion blur and exposure. Exposure affects overall image noise. Exposure and ISO affect image lightness.

You can set the camera to fully automatic, and not worry about any of this. You can take full control and manually adjust everything. You can pic and choose what you want to control and let the camera fill in the rest.

It turns out that with modern digital cameras, ISO is one of the things that you can let the camera decide. This is something that just wasn't available with film.

Allowing the camera to select ISO means that you don't have to hit a particular target exposure in order to get good image lightness. There's a wide range of exposures that will work.
 
The last couple of days in a discussion in the beginners forum, it had repeatedly been stated that getting an exposure with a digital camera isn't as critical as it was with shooting film.

One such quote ...

"With film, a "good exposure" was one that fell into the narrow range needed to produce a usable negative. If the exposure was too high, or too low, the negative wasn't usable (too thin, or too dense). With digital, the range of exposures that give a very usable result is well over 5 stops"

Today I decided to test this out. First I took in exposure with 0 EV on the dial...
You misunderstand the word "exposure". That's the amount of light that falls on the sensor. ISO is not a factor in exposure, however it is a factor in image lightness.

Try repeating the test, but this time don't pretend you are using a film camera - accept that you are now shooting with a digital camera. Your camera doesn't have any film in it, don't use a fixed ISO!

Try the following test:

Use a meter to determine a reasonable f/stop and shutter speed for ISO 100.

Put your digital camera in manual mode and select the meter recommended shutter speed and f/stop.

Remembering that this is not a film camera, don't select a fixed ISO. Set the ISO to AUTO!

Now take a test shot. Now reduce the "exposure" (light reaching the sensor) by one stop (faster shutter speed or smaller aperture). Take another photo.

Reduce by another stop and repeat.

Take 5 or 6 test shots, each with one stop less light than the previous shot.

What you should find is that your digital camera gives you quite usable images with good image lightness across that 5 stop exposure range.
I'm thinking that the results must be highly camera-specific. My Sigma does not offer auto ISO when the exposure mode is set to manual, so that's not even an option. Even if it was, I wouldn't use it since, unless I'm converting to black and white, exposing for ISO 200 compromises image quality more than I'm willing to live with, so this camera stays at ISO 100 whenever I want color output.

I've tested my Samsung enough to know that ISO 200 (its lowest setting) gives less highlight dynamic range than ISO 400 and above, and I also know that noise becomes unacceptable to me at ISO 800; therefore, this camera is always at ISO 400. Auto ISO simply lets the camera choose non-optimal settings.

Auto ISO also ties you to the built-in meter, and I often prefer using a handheld meter or the fun of relying on exposure charts. And unless you want to let the camera determine the brightness of the images you're producing, you have to resort to additional things like exposure lock or exposure compensation. Manual control makes it much easier for me to use a variety of exposure setting aids and to get exactly what I want in my out-of-camera images.
 
I'm thinking that the results must be highly camera-specific. My Sigma does not offer auto ISO when the exposure mode is set to manual, so that's not even an option. Even if it was, I wouldn't use it since, unless I'm converting to black and white, exposing for ISO 200 compromises image quality more than I'm willing to live with, so this camera stays at ISO 100 whenever I want color output.

I've tested my Samsung enough to know that ISO 200 (its lowest setting) gives less highlight dynamic range than ISO 400 and above, and I also know that noise becomes unacceptable to me at ISO 800; therefore, this camera is always at ISO 400. Auto ISO simply lets the camera choose non-optimal settings.

Auto ISO also ties you to the built-in meter, and I often prefer using a handheld meter or the fun of relying on exposure charts. And unless you want to let the camera determine the brightness of the images you're producing, you have to resort to additional things like exposure lock or exposure compensation. Manual control makes it much easier for me to use a variety of exposure setting aids and to get exactly what I want in my out-of-camera images.
Sounds like you have made some decisions that are reasonable for your situation.

The hard part is that there isn't always enough light to get the quality you want.
  • You may need to sacrifice desired depth of field to use a wider aperture.
  • You may need to accept some unwanted motion blur and use a slower shutter speed.
  • You may need to accept some unwanted visible noise by using a lower exposure (and likely a higher ISO).
In your situation, with your camera, you have made the judgement call that you care most about noise. That's OK. In your situation you likely don't want to use Auto-ISO because you want to target a particular exposure level.

Those more concerned with depth of field, probably don't want to use an auto aperture mode.

Those most concerned with motion blur, probably don't want the camera to chose the shutter speed.

These are all reasonable choices.

However, I think those who don't consider the possibility of allowing the camera to determine the ISO, are needlessly limiting themselves.
 
On one side of this debate we have the engineers of Sony, Nikon, Canon et. al, graduates of digital imaging programs at prestigious universities who represent the collective wisdom of more than 40 years of digital camera evolution.

On the other side we have the ETTR/ISO Invariance people who have come up with their own theories about how digital cameras work.

The engineers say "to get the most out of your camera, keep your exposures within the parameters based on the latest science and confirmed through extensive testing..."

The ETTR/ISO Invariants say "don't listen to the engineers, histograms are inaccurate, the 'right' exposure is the wrong exposure; for the best results, shoot your camera in ways it was never meant to be used."

Who to believe?
Not quite.


The camera engineers are saying "Use these settings and you will easily get reasonable out-of-camera JPEGS in most situations".

The ETTR people are saying "If you are willing to put a little more effort into it, you can squeeze some additional performance out of the system".

The ISO invariance people are suggesting digital ISO and film speed have vastly different underlying mechanisms, and therefore there are differences between the two.

Each are discussing different issues, and there are no inherent conflicts in their statements.
 
You're right--in fact, even in the +2 there are patches in the sky that were blown out. I did not take a +3 of the scene for this reason.

Here is an example with more room to decrease brightness in post. Note that the lava is blown out is each image except for the -2.67 EV.

Original images were taken at -2.67, -0.67, and +2.67 EV:

Exposed at -2.67, -0.67, and +2.67 EV
Exposed at -2.67, -0.67, and +2.67 EV

Here are corrected photos (brightened to match the middle photo at -0.67 EV):

Images above corrected to the same brightness as the middle image, -0.67 EV (+2 stops on the top, -3 stops on the bottom).
Images above corrected to the same brightness as the middle image, -0.67 EV (+2 stops on the top, -3 stops on the bottom).

With the Nikon D750 and D850--maybe all digital cameras?--there is more room to bring back details from an underexposed image. This may just be a function of the camera's metering doing a fairly good job in many situations of preserving highlights.

I often use this characteristic for a poor-man's fast lens. When I need a fast shutter speed and do not want overly high ISOs, I underexpose the image and brighten in post. Many on DPReview disagree that this is the best way to go about things--and I admit that they understand this stuff better than I do and are probably right. But in my mind it ends up looking better to brighten an image in post by 3 stops than to increase ISO from 1600 to 12,500, e.g. I have not really tested this--but I've mostly been happy brightening it in post so I didn't see a need.

FYI, Here is some detail with the bright lava and the dark surrounding area. I did not take any of these images to test exposure and used evaluative metering. This is probably why the corrected top image is brighter than the reference image. Also, with the different exposure lengths, the lava detail and smoke/steam look different.

Detail from the corrected images. Lava was clearly blown out in brightest image.
Detail from the corrected images. Lava was clearly blown out in brightest image.
 
I'm thinking that the results must be highly camera-specific. My Sigma does not offer auto ISO when the exposure mode is set to manual, so that's not even an option. Even if it was, I wouldn't use it since, unless I'm converting to black and white, exposing for ISO 200 compromises image quality more than I'm willing to live with, so this camera stays at ISO 100 whenever I want color output.

I've tested my Samsung enough to know that ISO 200 (its lowest setting) gives less highlight dynamic range than ISO 400 and above, and I also know that noise becomes unacceptable to me at ISO 800; therefore, this camera is always at ISO 400. Auto ISO simply lets the camera choose non-optimal settings.

Auto ISO also ties you to the built-in meter, and I often prefer using a handheld meter or the fun of relying on exposure charts. And unless you want to let the camera determine the brightness of the images you're producing, you have to resort to additional things like exposure lock or exposure compensation. Manual control makes it much easier for me to use a variety of exposure setting aids and to get exactly what I want in my out-of-camera images.
Sounds like you have made some decisions that are reasonable for your situation.

The hard part is that there isn't always enough light to get the quality you want.
  • You may need to sacrifice desired depth of field to use a wider aperture.
  • You may need to accept some unwanted motion blur and use a slower shutter speed.
  • You may need to accept some unwanted visible noise by using a lower exposure (and likely a higher ISO).
You've left out what seems to me the most obvious choice, and that is to simply put the camera away when there isn't enough light. I'm not a "get a picture at any cost" type of person, and I've always found enough picture-taking opportunities in good light to make the hobby satisfying. I also don't mind putting on a simple bounce flash unit for indoor snapshots (preferring that over direct flash).
In your situation, with your camera, you have made the judgement call that you care most about noise.
I want what I perceive to be the best combination of extended highlight dynamic range and low noise. On several cameras I've owned, the lowest ISO, while giving the lowest noise, also reduced highlight dynamic range (this includes models from Samsung, Olympus, Sigma, and Fuji), and I've always chosen to compromise noise a bit to gain more dynamic range. With most of the cameras I've owned, raising ISO even one more stop above that introduces noise compromises I don't want to live with, so there's usually been just one ISO I want to shoot at with any given camera.

When I'm out shooting, depth of field is the thing I care most about, preferring that everything in a picture looks reasonably sharp. I don't think about ISO when I'm out shooting because I've already chosen the ISO I like best. Like the prevailing light conditions, ISO is a condition that informs my shutter speed and aperture choices, not a variable that I want to change from shot-to-shot.
That's OK. In your situation you likely don't want to use Auto-ISO because you want to target a particular exposure level.
Perhaps you see it that way, but to my mind, I'm simply choosing an ISO to get image quality I'll be happy with. I did the same with film, happily using slide films from ISO 25 to 100, but never finding one above ISO 100 that looked good to me. I didn't think of it as targeting a specific exposure level; I was simply choosing film that gave results I liked and living with the limitations of its speed. It's very similar with digital cameras. I choose to use what I perceive to be the optimum ISO for image quality, but that has varied from model to model--on one camera it is ISO 100, on another ISO 200, and on another ISO 400.

I have had at least one digital camera that looked equally good to me across a range of ISO settings, an Olympus DSLR that looked fine to me whether at ISO 200 or 400. I have a feeling that some of the newer cameras might offer an even broader range of ISO's that I would find highly satisfactory. I can see where that would open up ISO selection as a field variable rather than being a fixed condition for exposure choices.
Those more concerned with depth of field, probably don't want to use an auto aperture mode.

Those most concerned with motion blur, probably don't want the camera to chose the shutter speed.

These are all reasonable choices.
I agree.
However, I think those who don't consider the possibility of allowing the camera to determine the ISO, are needlessly limiting themselves.
When you use auto ISO, do you also use exposure lock and/or exposure compensation to control the out-of-camera image brightness, or are you content that things will be close enough and that you can do fine adjustments later with an image editor? Part of my satisfaction in the hobby is getting out-of-camera images that look the way I want them to without additional manipulation, and while exposure lock and exposure compensation might work, I find simply setting everything manually to be simpler and easier.
 
You've left out what seems to me the most obvious choice, and that is to simply put the camera away when there isn't enough light. I'm not a "get a picture at any cost" type of person, and I've always found enough picture-taking opportunities in good light to make the hobby satisfying. I also don't mind putting on a simple bounce flash unit for indoor snapshots (preferring that over direct flash).
There's nothing wrong with your choices. However, my clients can get annoyed if I don't get the shot, and lighting is not always ideal.

In your situation, with your camera, you have made the judgement call that you care most about noise.
I want what I perceive to be the best combination of extended highlight dynamic range and low noise. On several cameras I've owned, the lowest ISO, while giving the lowest noise, also reduced highlight dynamic range (this includes models from Samsung, Olympus, Sigma, and Fuji), and I've always chosen to compromise noise a bit to gain more dynamic range. With most of the cameras I've owned, raising ISO even one more stop above that introduces noise compromises I don't want to live with, so there's usually been just one ISO I want to shoot at with any given camera.

When I'm out shooting, depth of field is the thing I care most about, preferring that everything in a picture looks reasonably sharp. I don't think about ISO when I'm out shooting because I've already chosen the ISO I like best. Like the prevailing light conditions, ISO is a condition that informs my shutter speed and aperture choices, not a variable that I want to change from shot-to-shot.
Essentially, you are balancing three things; image noise, depth of field, and motion blur.

You seem to be saying that image noise (and dynamic range) are most important to you. That's reasonable. But not everyone has the same priorities.

That's OK. In your situation you likely don't want to use Auto-ISO because you want to target a particular exposure level.
Perhaps you see it that way, but to my mind, I'm simply choosing an ISO to get image quality I'll be happy with. I did the same with film, happily using slide films from ISO 25 to 100, but never finding one above ISO 100 that looked good to me. I didn't think of it as targeting a specific exposure level; I was simply choosing film that gave results I liked and living with the limitations of its speed. It's very similar with digital cameras. I choose to use what I perceive to be the optimum ISO for image quality, but that has varied from model to model--on one camera it is ISO 100, on another ISO 200, and on another ISO 400.
Yes, you didn't think of it as "choosing exposure," but's that what you were doing when you chose a particular speed film.
I have had at least one digital camera that looked equally good to me across a range of ISO settings, an Olympus DSLR that looked fine to me whether at ISO 200 or 400. I have a feeling that some of the newer cameras might offer an even broader range of ISO's that I would find highly satisfactory. I can see where that would open up ISO selection as a field variable rather than being a fixed condition for exposure choices.
Of course the images look different at different exposures. Change the aperture you get different depth of field, change the shutter speed you get different motion blur, change the exposure you get different image noise.

A photographer has the option of choosing how he wants to balance these.

My point is that with film, exposure (light hitting the film) had to be treated specially. You had to match it to the film you were using. Therefore we learned to treat it differently than aperture and shutter speed. You had to start with a target exposure (film speed), and then figure out aperture and shutter to match.

Digital lifts that restriction, you have the option of starting with depth of field or shutter speed. You can even treat all three equally.
Those more concerned with depth of field, probably don't want to use an auto aperture mode.

Those most concerned with motion blur, probably don't want the camera to chose the shutter speed.

These are all reasonable choices.
I agree.
However, I think those who don't consider the possibility of allowing the camera to determine the ISO, are needlessly limiting themselves.
When you use auto ISO, do you also use exposure lock and/or exposure compensation to control the out-of-camera image brightness, or are you content that things will be close enough and that you can do fine adjustments later with an image editor? Part of my satisfaction in the hobby is getting out-of-camera images that look the way I want them to without additional manipulation, and while exposure lock and exposure compensation might work, I find simply setting everything manually to be simpler and easier.
My camera allows me to use exposure compensation in various auto modes. In manual mode you control brightness by manually setting ISO. However, unlike film, you can set ISO last, you don't have to set it first.

I typically get good quality JPEGs right out of the camera. However, I find that I frequently can get better images by working from the RAW file. For what I charge clients, they deserve the better image.
 
If the entire image is clipped highlights (or shadows), there is no data to recover, hence the white image just turning gray. That's not really a fair example.

Here are some images showing an almost 5-stop processing range. There is a small amount of clipping on each end, but not enough to make a visual impact (so it is not quite a full 5 stops).

The first panel shows three images: -3 EV, 0 EV, and +2EV. These are RAW files processed in LR with "daylight" WB and no other adjustments. (All images: Nikon D850, Sigma 20mm f/1.4 taken within 70 seconds of each other at f/16, ISO 64, and 1/100, 1/13, and 1/3 seconds.)

View attachment cf6132d5b78e4fe6998b51c4aaa6586d.jpg
Exposure compensation: -3 EV, none, and +2 EV (right).

The next panel shows the same images with the exposure corrected (+3 EV, 0, -2 EV):

View attachment ee0c0ee177f94b25b5c4c2f0b71dfdac.jpg
Each image adjusted in LR: +3 EV, 0, and -2 EV (right).

Looking at them in full size, there are some clear (but minor) differences in color, but they are pretty similar overall.

I agree with your main point: watch for clipped highlights--no recovery. But there is a lot of data hiding in the shadows!
Exactly. This is true of colour slide film too.

The problem with digital is that the best exposure for a JPG may not be the best for raw. ETTR is really an approach for raw files, and is liable to give over bright JPGs in some lights.
 
. . . Yes, you didn't think of it as "choosing exposure," but's that what you were doing when you chose a particular speed film. . .
Thanks for the lengthy reply, lot's of good stuff there. I've got to run out the door momentarily and won't see this thread again for several hours, but the quote above still seems odd to me. It sounds like you're saying that for any particular ISO setting, there is only one exposure choice (in terms of light hitting the sensor, not the various combinations of shutter speed and aperture that will yield that same amount of light). But don't people still choose what level of exposure to give a particular image? For example, choosing to let a back-lit subject be silhouetted vs. giving enough exposure to show more detail? A person might choose to expose a bit more than the meter reading to emphasize how bright and hot a desert scene is, or underexpose on an overcast day to reflect that it was a dark day. I'll often underexpose images taken under the trees to help differentiate them from those taken under open sky or direct sunlight because it seems more natural when viewing a "slide show" that those under the trees pictures should be a bit darker. Even with a fixed ISO, I'm still making exposure choices, and quite possibly different choices than a person with me who's using the same ISO setting.
 
. . . Yes, you didn't think of it as "choosing exposure," but's that what you were doing when you chose a particular speed film. . .
Thanks for the lengthy reply, lot's of good stuff there. I've got to run out the door momentarily and won't see this thread again for several hours, but the quote above still seems odd to me. It sounds like you're saying that for any particular ISO setting, there is only one exposure choice (in terms of light hitting the sensor, not the various combinations of shutter speed and aperture that will yield that same amount of light). But don't people still choose what level of exposure to give a particular image? For example, choosing to let a back-lit subject be silhouetted vs. giving enough exposure to show more detail? A person might choose to expose a bit more than the meter reading to emphasize how bright and hot a desert scene is, or underexpose on an overcast day to reflect that it was a dark day. I'll often underexpose images taken under the trees to help differentiate them from those taken under open sky or direct sunlight because it seems more natural when viewing a "slide show" that those under the trees pictures should be a bit darker. Even with a fixed ISO, I'm still making exposure choices, and quite possibly different choices than a person with me who's using the same ISO setting.
With film there's a response curve - the density you get for a particular amount of light hitting the film. Increase/decrease the overall exposure of your image image and you hit a different part of the curve. If that part has a noticeably different slope, then you get a noticeably different image. Where the slope is steep you get good contrast. Where the slope is shallow, you lose contrast. Where the curve is flat, you don't get any detail.

The speed rating of the film has to do with the overall recommended exposure for that film. In other words how to hit the sweet spot on the curve for a typical scene. Clearly you can deviate from that for artistic purposes.

In terms of overly bright, overly dark, backlit or atypical subjects; yes, you probably want to exercise some artistic judgement to determine where on that curve you want various parts of the scene to fall. If the subject is backlit, you may want the subject to fall into the curve's sweet spot, and allow the background to flattened out density. Artistically, you could also choose to have the background hit the sweet spot, and lose the subject to minimum density.

If you were shooting a white wall, you might want to allow the wall to fall into the sweet spot (typically where grey falls), and lighten the image in the darkroom. This would get you better details in the wall.

No matter how you choose to expose, that "sweet spot" of exposure was determined when you loaded the film into the camera. You're stuck with that same sweet spot of exposure for the entire roll.

Digital doesn't work that way. The ISO setting of the camera determines the sweet spot for the camera-produced JPEG. If you are shooting RAW, you have a lot of leeway to deviate from that.

Most importantly, the ISO isn't fixed for a series of shots. Like aperture an shutter speed, you can change ISO on a shot to shot basis. Like aperture and shutter speed, many cameras allow you to use "Auto-ISO", where the camera will find the sweet spot that matches your aperture and shutter speed. As with other Auto modes, many cameras will allow you to use "Exposure Compensation" to bias the camera's metering system.

Exposure Compensation is very helpful any time you are relying on the camera's metering system, and you are unhappy with the camera's conclusions.

====

Let's go back to the scenario of a shooting a backlit kid running at the beach. Suppose you want two shots, one with the kid in silhouette with a nicely detailed sky behind her, and one where the kid is nicely detailed, but the sky is blown out.

With digital you can pick the aperture that gives you just the depth of field you want, and then the shutter speed that gives you just the right amount of motion blur. You can then take shots at various ISO values to get the image you want.

With film, you can't use the same aperture and shutter speeds for both images, as you need to hit the target exposure for the film. Thus you need to compromise on either motion blur or aperture.

Of course, you could shoot the digital camera the same way as a film camera, but you don't have to. With film you even have the option of spreading the changes around. You can make smaller changes to each of all three (aperture, shutter and ISO).

Obviously, not all photography is as challenging as a backlit kid running along the beach. However the choices that digital makes available in that scenario are also available in less challenging situations.
 

Keyboard shortcuts

Back
Top