Is the camera meter always right? And do you believe it?

[KVH]

So. Way back when terrible dinosaurs walked the earth, there was this thing called film. In those wild and wooly days, before the second coming (of Photoshop), you not only had the ability to manipulate the exposure in the camera (time, aperture, ASA, flash), but you had two more chances after the shutter fired. A fairly wide variety of chemical types, processes, temperatures, and times. Then you could pick from another wide variety of papers, chemicals, yet more exposure controls, filters, yadda, yadda, yadda. With hundreds of types of film, the combinations were endless.

Jeez, doesn't this start off like a typical disney-oriented "I miss the old days" post? Anyway. Short story long.

Common wisdom was you overexposed negative film (up to 2 stops) and underexposed positive (0.5-1 stop). My question (finally) is whether you just pay attention to what is the "correct" exposure in your digital work and try to correct everything in PP? Or do you go beyond and use compensation in most of your shots hoping to save time later?

NOTE: I do not include HDR in this. That mess is a whole 'nother thread.

 

[stewart715]

The metering mode will drastically change what the meter says. What you want it to meter and what it is metering are many times two different things. That being said, with digital and shooting RAW, the metering is usually close enough that you can tweak it to your liking. I still do find myself over exposing by a stop often though because it's usually easier to remove a little light than add it (at least in my experience).

 

[pjacobi]

Light meters are calibrated to 18% grey. Certain subjects will always fool a light meter into under or over exposure, such as snow or a black tuxedo. For these special subjects you need to meter off of a "grey card" for correct exposure. Some photographers have inserted a grey card inside the lens cap for easy access.

-Paul

https://en.wikipedia.org/wiki/Gray_card

 

[bob100]

So. Way back when terrible dinosaurs walked the earth, there was this thing called film. In those wild and wooly days, before the second coming (of Photoshop), you not only had the ability to manipulate the exposure in the camera (time, aperture, ASA, flash), but you had two more chances after the shutter fired. A fairly wide variety of chemical types, processes, temperatures, and times. Then you could pick from another wide variety of papers, chemicals, yet more exposure controls, filters, yadda, yadda, yadda. With hundreds of types of film, the combinations were endless.
Jeez, doesn't this start off like a typical disney-oriented "I miss the old days" post? Anyway. Short story long.
Common wisdom was you overexposed negative film (up to 2 stops) and underexposed positive (0.5-1 stop). My question (finally) is whether you just pay attention to what is the "correct" exposure in your digital work and try to correct everything in PP? Or do you go beyond and use compensation in most of your shots hoping to save time later?
NOTE: I do not include HDR in this. That mess is a whole 'nother thread.

short story short - light meter in the camera and digital photography makes it easy
choosing the "best" metering mode and shooting raw also helps with the "correct" exposure
www.flickr.com/photos/mmirrorless

 

[KVH]

Ah. So what I'm hearing is y'all spend a lot of time chimping and in PP? Bracketing doesn't seem to be a thing? Probably great for landscapes and portraits (if your model doesn't mind waiting around). But what do you do for grab shots and things that happen in the moment and are then gone? What would you do if you couldn't do shot review?

I'm not judging, just curious how others shoot some things w/o setting compensation. I mean technically, there's a 50% any given shot will be under-exposed. And if you never capture the light in the first place, PP ain't gonna be able to add it back in later. And I hate spending time doing PP (but I've lost count of my darkroom hours (yeah, I know)).

 

[jec6613]

Well, I shoot Nikon, so I'm most familiar with their meters, but each manufacturer has their own special sauce.

With the 3D color matrix meter, used on the F5, F6, and almost all digital bodies, it will favor underexposure. This is because in post processing digital exposes like positive film: if you blow a highlight, it's gone, but it can recover upwards of 5-6 stops from the shadows with newer sensors. It also attempts to use skin tones and meter to those using an image lookup table, but only if the human subject is large enough on the meter - so portraits only on the 512 segment meters, but the newer ones in the D5, D500, D7500 and D850 can do full face recognition and AE. And, again, if there's too much contrast: favor underexposure.

It should be noted that unless you spot meter or CW, Nikon will not shoot for 18% grey, it uses color information and has done so for ages with its matrix meter. With spot and CW though, yes, it will do a basic panchromatic-esque meter (though the meter still sees in color so can get a bit fooled by it).

In some newer bodies there's also a highlight weighted metering option, which is amazing for lit stage performers on dark backgrounds. It attempts to expose the highlight areas properly and lets the rest go dark: perfect for cleaning up in post processing while preserving the most raw data.

So, basically, what I do on my D500 is trust the meter to get, "Close enough," that the raw converter can handle it. On my Df with its more primitive meter, I will sometimes dial in a bit of negative EC just in case if it's contrasty out, which works great since the sensor (and all other modern Nikon sensors except the 20MP ones) are ISO invariant: I can easily fix the ISO in post with a slider, and within 2-4 stops it's indistinguishable from getting it right the first time. On, and for really high contrast static subject? Yeah, I shoot a 5 shot bracket at -4, -2, 0, +2 and +4 stops. On a D500, that takes half a second.

One other thing: if you check your camera histogram to verify exposure, it's using the embedded JPEG settings. If you want a histogram closer to what the sensor actually captured, use a toned down setting: Flat or Neutral in Nikon-land.

 

[zackiedawg]

I prefer to set up my exposure and entire shot in the camera when taking the shot, to avoid the need to post-process, so I use the meter as a tool to kick off the settings for the shot, adjust those via the EV as needed depending on the scene, and then adjust based on my own personal experience and knowledge from film days to adjust the exposure for how I intend to take the shot, ie: if I wanted to intentionally overexpose the meter to go for a high-key look or capture shadows, or underexpose to shoot a silhouette or partial silhouette, and so on. As a film photographer who moved into digital, it's one of the things I enjoyed most about digital photography - being able to 'see' my shot before actually taking it and making adjustments on the fly. As long as the exposure is simulated in live view or EVF with settings applied, it's easy to apply artistic adjustments in camera rather than having to process for them later...if that's something you like to do.

The meter may be 'right' in achieving a 'proper' exposure for a particular scene, but not necessarily in achieving how I wish that scene to be captured...the 'right' exposure isn't always what the photographer may be going for. Low key, high key, silhouette, HDR, B&W, etc all require adjustments from a 'correct' meter. If wishing to capture the photograph when taking the photo exactly how you intend to display it, without intending any RAW processing or post processing, one needs to apply that intended bias to the exposure when taking the shot. If intending to do post-processing or RAW conversion, then you instead shoot to maximize shadow or highlight range without blowing either out beyond recovery, since all of the processing will be done later with the computer.

 

[jec6613]

While I agree that getting it the way you want it to look in camera is ideal, and I certainly do it with film when I shoot it, in practice I've found that what I really want is to capture the best and most data possible on the sensor. This is because, unlike with film, I can go in and change contrast, saturation, sharpening, exposure curve, and color balance after the fact to make it look the way I remember the scene, or to tell a story. If I do a proper exposure, or hit high key/low/key or whatever in camera, I'm often losing quite a bit of dynamic range and clipping highlights, which I could otherwise use to provide additional detail - and I can still make the highlights blow if I want. I can even apply the exposure curve of a film, if I want - I have about 20 stocks loaded into my editing program. I can't make many of those changes in camera.

At low ISO modern sensors capture 12+ stops of dynamic range accurately, so I easily have 4 extra stops over even negative film, let alone slide film, and because it's digital I can change those things mathematically perfectly.

 

[bob100]

I think we can all agree, the technology that put a man on the moon has (eventually) created amazing small cheap digital cameras that can fit in your pocket

www.flickr.com/photos/mmirrorless


Untitled by c w, on Flickr