What "Exposure Compensation" does

M Mode with Auto ISO is often used for wildlife photography. Of course, that mode is not pure M mode.

Can you elaborate on which cameras you are aware of that change metering based on the WB setting? I haven't seen that behavior with any Nikon cameras. Also, I would suggest that if that is the case, then switching to UniWB should generally be done in conjunction with Manual Exposure mode or applied with an understanding of how it affects the exposure and with appropriate EC applied.

One other thing here regarding EC, with Nikon's Matrix metering EC can be a prescription for over and under exposures because the "matrix" sometimes adjusts the EV by as much as a stop or more when making relatively subtle changes in what's being focused on, the percentage of sky in the frame, etcetera. Trying to determine an appropriate EC when there are so many variables that will do their own exposure compensations is like trying to steer a boat when you don't know which direction the next wave is coming from. Therefore, I always use Center-weighted metering (set to "Average") on my Nikon cameras because that takes a lot of the guesswork out of what the metering for the scene will be.

I agree but iso does have an effect on the raw file lightness.

Touche.

You have already pointed out that the count that's stored in the raw file is what's being enlarged, extended, added or increased. But you have said that beginners shouldn't have to worry about those counts, or DNs or whatever you want to call them.

I sorta of get this, but it's hard to explain raw clipping and how to look at raw files to see what's going on with your images without worrying about DNs.

Nikon MILC cameras meter on the basis of the JPEG preview image. Anything that affects the WB of that image will affect the metering. Is it a significant effect with normal WB manual variations? I don't know.

One exception to that would be Auto-ISO, at least on the Nikon cameras that I have used. Then the Raw file is effected because changing the ISO setting changes the Raw file.

That is unfortunate behavior, I think, of your camera's metering system. I often like to see what is happening to the metering if it was making automatic changes to the exposure even though I'm shooting in Manual Exposure mode. Nonetheless, EC by itself doesn't have an effect on the Raw file, it only effects the Raw file when it changes how the scene is metered (whether that is by you responding to it, or the camera responding to it),

Sure, but that's besides the point.

Again you make a valid point, but there are times when I don't want the EV changing in response to changes in the background or where my camera's focus point is, so then I'm using Manual Exposure mode and the EC may be initially set to a value and will be recommending all sorts of adjustments to my selected EV that I ignore and do not effect the Raw file.

Okay, but I would like to see it tested to see if the metering can be changed by changing the WB setting. I'm guessing that the camera determines the appropriate WB setting and then sets the EV based on that, and maybe that happens with Matrix metering on Nikon cameras in general (I don't know). All I can say for certain is that I have tried to change the metering of various scenes by changing the WB setting on my Nikon DSLR cameras and have never seen it change in spite of slamming the WB setting from one extreme to the other.

I'm not sure about its effect on "lightness," but on my Nikon cameras it can cause overexposure.

I was talking about Nikon MILCs like the Z6, Z7 and Z9. I thought I was explicit about that. I'd be surprised if the WB setting on a Nikon DSLR affected exposure.

The Hasselblad X2D-100C works the same way. I consider that a drawback, since it impedes Zone System like tone placement. That's easily worked around, though, and you could argue -- and I have -- that that doesn't make sense for a raw shooter.

With all cameras, EC has no effect in M mode and M-ISO. What is the difference between Z7 and X2D?

I don't have the Z7 with me, but I remember cameras (but not which cameras) where the EC setting biased the meter reading that you see in M mode. I only mentioned the 'blad because that's what I can lay my hands on.

Define raw file lightness, please.

Now I understand what you meant. Z7, unlike X2D, shows the effect of EC in M mode in the meter reading.

EC should affect metering. There is no reason that EC can’t bias the meter in a fully manual mode. While it won’t directly change exposure, the meter can certainly influence the photographer to alter the settings. If the photographer is adjusting exposure to according to the meter, than biasing the meter will influence the photographer to pick a different exposure.

I started typing a much longer response but have only half finished it because I felt it was getting way too far into the weeds.

I fundamentally disagree with this statement.

Exposure compensation exists because light meters(as they exist in a camera) are fundamentally just measuring and reporting to you the light the meter sees reflected from the scene and telling you what exposure will render at least some elements at middle gray(18%) on a properly developed negative for a film with a given ISO. This is at least is true of more primitive metering types-averaging/center weighted averaging, partial/spot metering(to use Canon and Nikon terminology for the same basic thing) and true spot metering(a very small area of the frame, not the gigantic 12mm circle Nikon calls a spot meter).

In a scene with a lot of contrast, we have to pick what part of the scene we want to expose for "correctly". This can be especially important when we are working with a high contrast film like Velvia with 4-5 stops of usable dynamic range, or even certain print films like Tech Pan or TMZ(or in situations where we intentionally underexpose and over-develop-aka "push" film). There are also times where we don't want a subject rendered as middle gray, but rather darker or lighter to be consistent with the actual appearance of the subject. Ansel Adams developed and exhaustive characterized a frame work for thinking about exposure that is often referenced but seemed to rarely be fully understood or used called the zone system(in his seminal book The Negative). Thinking in zone system terms, an averaging meter gives us a reading that will render the scene, on average, in Zone V. Using the system properly, we use a spot meter and might choose to place our foliage in Zone V or caucasian skin in Zone VI(exposures that will give reasonable representations of how they actually appear).

If we are using an uncoupled meter, or a coupled TTL meter, in a camera with no exposure automation it's relatively easy to over-ride the meter. In a scene dominated by snow, we might choose to expose 3 stops higher than the meter indicates so that the snow is rendered as white and not gray. If a scene is strongly backlit, we might set our exposure higher than the meter says by a couple of stops because we know that the meter reading will leave our subject deeper into the shadows than we desire, or alternatively we might choose to use a meter reading from a different area similar to our subject and use that in our final composition. A prime example of the latter is a person in front of a daylight window, where taking the reading from an averaging meter as-is will typically place the person deep into the shadows(zone I-III maybe in zone system terms). We might point the meter down toward their feet to cut most or all of the window out of the frame, meter, and then set exposure. By the 70s, when coupled TTL metering without automation was the norm, and cameras like the Nikon F2 and Pentax K1000 used center weighted averaging or Canons(FTb and F-1) used partial metering, experienced photographers understood how the meters worked and could quickly point the camera, set the exposure(either taking the meter reading as-is or adjusting it) and then compose to the final image.

Automatic exposure took away our ability to just change the shutter speed or aperture on a whim to compensate for unusual light or artistic effect. All of the early AE cameras I can think of(Nikon FE, Canon EF and AE-1, Olympus OM-2) had center weighted average meters, and left to its own device the camera would just set the exposure to whatever the meter said. This could be over-ridden somewhat with judicious use of exposure lock(which most, but I don't think all, cameras had) but that still wasn't a perfect solution. Even worse, some AE cameras made manual exposure very clunky-outside the New F-1, Canon FD mount cameras that were capable of shutter priority did not have a way to display set aperture in the viewfinder, so you either had to carefully count clicks or just look at the aperture ring.

Exposure compensation came about as a quick and easy way to tell the meter "you're wrong and this is how you can fix it" in automatic exposure modes. I learned photography on a Canon A-1, which traveled the world with me and shot I don't know how many hundreds of rolls of film(I have it still but it's more or less retired, even though it was already into a second 25-year-old life when I bought it). I never actually used it in manual mode outside low-stakes playing with the camera and using an uncoupled auto flash like a Vivitar 283 because it's so clunky to use, but I did make judicious use of AE lock and of course dialing exposure compensation as appropriate.

Meters did start getting smarter. Nikon brought us the 5-segment Matrix meter in the 80s which would attempt to evaluate the scene, determine the subject, and apply its own exposure compensation. The FA, the first "AMP" meter camera, did a decent job but you still needed to have some idea what was going on and use AE-L or EC when the scene warranted, and in 2023 if you're trying to use an FA you also cross your fingers that the electronics won't flake out and give you a blank or underexposed frame when you release the shutter. Really, though, just don't use an FA now unless you trust it's actually 100% working...Canon(in the T90) and Olympus(in the OM-4) gave us the elegant if not slow multi-spot meter that, used correctly, let you basically use your own actual knowledge of the scene to do what Nikon's matrix tried to do automatically.

Now, matrix/multi-segment meters mostly just work. They have bunches of segments, a lot more computing power to analyze a scene, collect color information, and also weigh user input like the selected focus point. Most of the time now, I just compose, move the focus point where I want it, and "trust the Matrix." Just last night, I worked up about 400 RAW files I took over the weekend. My new-to-me D5 consistently amazed me that if I was happy with the composition/focus I just hit "P" in Lightroom and went on to the next one. I was shooting it along side my very familiar D850, and just like all my D8xx cameras(I started with a D800 in 2017, bought a D810 in 2020, and then the D850 last year) almost every "keeper" still needed a levels/curves tweak and often a white balance nudge for me to be happy. Still, though, they're pretty darn good to just leave the camera in aperture priority and bump ISO if my shutter speed goes too low for the situation. That's basically how I've shot digital since I decided it was worth supplementing my kit with a DSLR, with the only exceptions being when I'm working with studio strobes or when I have the time to do more than compose, focus, and make sure the shutter speed and aperture work for the situation and get the photo NOW.

I realize that's a lot of rambling, and to get around to it I find that when I'm using modern multi-segment meters I almost never touch exposure compensation. If I'm not happy with something I sort it out in post, as it's usually still a minor adjustment. Even though CWA and spot metering are close at hand on my cameras, I could probably count on one hand the number of times on my DSLRs I've actually found a need to use them.

attach one to the common feed to control all of the pixels at once 🙄 LOL

There isn't any place in a CMOS sensor with on-chip column ADCs -- the standard these says -- where all the pixels are multiplexed into a common line.

You present some good reasons for the need to bias the meter. I don't think anyone has questioned that.

Where we differ is whether the name "Exposure Compensation" is a good name.

Suppose you are shooting at f/5.6, 1/100 and ISO 200. Perhaps the resulting camera produced JPEG is lighter than what you want. Biasing the meter by a stop, may result in f/5.6, 1/100 and ISO 100. Same exposure, but a darker camera produced JPEG.

My point is that with a modern digital camera, "Exposure Compensation" may result in a change in ISO, and not a change in "exposure". Therefore the name "Exposure Compensation" is confusing.

"Meter Bias" or "Meter Compensation" would be a better name, as that tells us what the control does. Even "Lightness Compensation" is better, as it describes the desired result.

"Exposure Compensation" describes how a film camera reacts when you bias the meter. While a digital camera may react in that way, it is not the only way it can react.