Subject too dark...

[J-see]

I also look at A or S as a task taken up by the camera, freeing me from constantly worrying about changing light. They are assists not an encumbrance. If I do not agree with the camera's logic, I can always over rule it (with exposure compensation). To be irritated with the assistance the camera is offering you, is in mind a bit churlish, you asked for it and then you do not like what it suggests, after all it is trying to do its job. As others have commented, if you do not like what the camera dishes out to you, ignore it - go fully manual.

The whole manual = more control idea was true in the days of yore but now it is simply an old wives tale. Manual only makes sense when you control the light.

People talk about the exposure triangle but apparently few understand that this triangle has a fixed area which is defined by the available light. The three sides are Aperture, ISO and Shutter. When we change one, we have to adjust one or both since the area size has to remain identical for perfect exposure. Whatever value we use for any of the sides, if the others are not adjusted accordingly we change the total area and thus under or overexpose. Manual is adjusting all three sides until you have an ideal compromise that doesn't change the area. But, and this is the illogical part of manual, if you change two sides of any triangle and you don't want to change the area, there's only one option for the third.

Guess what the cam picks? Like I said, the difference is either me selecting a number or letting the cam select that same number.

Btw, going manual won't change the fact that the cam warns about the scene being underexposed when it's still brighter than reality.

 

[Horoscope Fish]

Have to agree with @J-see on this one.

I'm completely comfortable shooting in full Manual, but it seems like a pointless exercise most of the time. As the photographer I'm *always* in complete control of my exposure; semi-automatic modes like aperture and shutter-priority allow the camera to "suggest" certain shooting parameters to me, but they don't wrest control away from me.

A big part of the problem, if you can even call it that, I think, is people forget the light meter in their camera is designed to expose for 18% grey 100% of the time. Further, while a lot of photographers seem to understand that concept on an intellectual level, they have not grasped the full implication of what that means and how to compensate for it. Learning to "see" in 18% grey is a skill every photographer should strive to master, in my opinion.

....

 

[J-see]

A big part of the problem, if you can even call it that, I think, is people forget the light meter in their camera is designed to expose for 18% grey 100% of the time. Further, while a lot of photographers seem to understand that concept on an intellectual level, they have not grasped the full implication of what that means and how to compensate for it. Learning to "see" in 18% grey is a skill every photographer should strive to master, in my opinion.....

The problem is that light metering apparently is not identical for all cams and thus neither is perfect exposure. I read about some noise comparison between a couple of Nikons a while ago and while they all used the same settings and lenses, some were consistently brighter than others.

 

[Horoscope Fish]

The problem is that light metering apparently is not identical for all cams and thus neither is perfect exposure. I read about some noise comparison between a couple of Nikons a while ago and while they all used the same settings and lenses, some were consistently brighter than others.

Well, let's be careful... What exactly is "perfect exposure"? What amounts to "perfect exposure" for one scene, obviously, is not going to be "perfect exposure" for another.

In my opinion if my light meter is NOT consistently exposing an 18% grey card to 18% grey then there's a problem. Now, if it is *consistently* exposing to, say, +.5EV, I can deal with that in the Fine Tune Exposure setting (at least on my D7100). Same goes for a consistent -EV; as long as it's consistent, it can be dealt with. In my opinion it's still a defect but a defect that can be easily worked around. If the exposure was INconsistent by any significant degree then the camera is simply defective and needs to be repaired.

My point in all of this was that meters give us a BASELINE of exposure; not "perfect" or even "correct" exposure (except by coincidence); correct exposure is always up to the photographer to decide. Meters meter for 18% grey 100% of the time; that is the baseline we, as the photographer must work off of.

....

 

[J-see]

Perfect exposure is probably an ideal that won't be possible in reality.

Thinking about it, it's not necessarily light metering. Anyone with two different cams can test it, same settings, same lens, same shot. The RAW files should be close to identical. If not, either light metering differs and 18% grey means something else for each cam or the translation of what is captured is different which means, each different sensor has a different output. The problem is that if different cams have a differently exposed shot, only one of them is correct (or none). The rest is either slightly over or underexposed.

Now even when, that's not what I have issues with. Slightly too bright seldom is a problem. It needs to overexpose quite some to have white noise appear which can no longer be corrected in post. Underexposing is something else; dark noise appears rather quick and doesn't disappear when correcting exposure in post. Which is why I only compensate exposure to brighten.

But the main issue is that when I try to shoot at the "end", it warns about underexposure while the shot is perfectly reasonable (for the conditions). It's too early and I can't rely on the cam. The only option left is take a shot, check, correct or not and try again. Rinse and repeat. That annoys me but I learned to live with it. Yet I'd be a happy bunny if I could adjust the warning levels until they match the "real" limit.

 

[PaulPosition]

What's the camera supposed to do? Second guess you and check the clock "oh it's night!" ?

It's just telling you "hey, that last stop of shutter speed, I can't compensate for it anymore as I ran out of aperture to bring the (spot/center average/matrix-mess) near 18% like you're asking me to do". No message would make no sense, the camera has to warn that it can't do what's expected of it (ie, in S mode, to adjust aperture).

 

[J-see]

What's the camera supposed to do? Second guess you and check the clock "oh it's night!" ?

It's just telling you "hey, that last stop of shutter speed, I can't compensate for it anymore as I ran out of aperture to bring the (spot/center average/matrix-mess) near 18% like you're asking me to do". No message would make no sense, the camera has to warn that it can't do what's expected of it (ie, in S mode, to adjust aperture).

The cam doesn't need to tell me anything. Like I said, if I could set the limits, I'd be a happy bunny. If I could disable the warnings, I'd be satisfied too yet if it's an option, they sure did hide it well.

 

[wornish]

All cameras are made for the average user, if thats not you then simply ignore the warning messages - problem solved.

 

[J-see]

All cameras are made for the average user, if thats not you then simply ignore the warning messages - problem solved.

And that's what I'm doing since a while.

It's just that people keep trying to convince me only amateurs shoot A or S.

 

[aroy]

One method that many professionals use to get consistent exposure is to shoot a standard colour chart (which should have grey scale also), before, after and midway through the session. They then use software to normalize all the intermediate shots and correct the colours to be consistent with the chart. That way they come very close to repeatable exposure and colours.

I could also use that technique, but in my case where shots are random and I am too lazy. Maybe if the images are for commercial purpose and colour accuracy is of paramount importance, I would do it, other wise I like what comes out.

 

[J-see]

I was bored so I did some testing.

Don't check the quality of shots, it's about the exposure.





The first shot is normal exposure. The second is compensated to -2EV. The second is accurate (as reality) which implies when light is low, the D3300 tends to overexpose by two stops. It isn't even that bright when I'm in the room and the bag is black, not gray.

It's the same using center or matrix metering.

Tomorrow I'll check if there's a difference during daylight. I somehow hope it does the same when plenty of light.

 

[Horoscope Fish]

The first shot is normal exposure. The second is compensated to -2EV. The second is accurate (as reality) which implies when light is low, the D3300 tends to overexpose by two stops.

I'm not following your logic.

The meter doesn't expose to make things look like our eye's see them, it exposes for middle gray (18%)... ALL the time.

....

 

[Marcel]

I was bored so I did some testing.

Don't check the quality of shots, it's about the exposure.

View attachment 119394

View attachment 119395

The first shot is normal exposure. The second is compensated to -2EV. The second is accurate (as reality) which implies when light is low, the D3300 tends to overexpose by two stops. It isn't even that bright when I'm in the room and the bag is black, not gray.

It's the same using center or matrix metering.

Tomorrow I'll check if there's a difference during daylight. I somehow hope it does the same when plenty of light.

If you meter something black, it will always over-expose. If you meter something white, it will always under-expose. The meter is trying to make everything 18% grey. It's how it's supposed to work. The camera has nothing to do with it, but the human taking the reading has.

 

[J-see]

Why then is matrix metering exactly identical as center?

 

[Marcel]

Why then is matrix metering exactly identical as center?

It would probably depend on the nature of your subjects. I really much doubt that it would always be identical.

Just try to use spot metering one of these days and you'll see how different it can be from matrix.

 

[Horoscope Fish]

Why then is matrix metering exactly identical as center?

Because, as I said, the meter is trying to expose to the same standard of 18% grey... ALL the time.

How you meter the scene -- using Spot, Center Weighted or Matrix -- doesn't matter because the meter does one thing and one thing only: it attempts to expose the scene at 18% gray; and that's regardless of "how" you meter that particular scene. It's just that most things look properly exposed at 18%. Generally. So you can insure that your subject is pretty much properly exposed by using "Spot" and keeping them at precisely 18%.

What you meter on can change exposure, as Marcel points out, but the scene will *still* be exposed to the same 18% gray overall. If you meter something black it will "over expose" to bring the scene to (18%) gray. If you meter something white it will "under expose" to bring the scene (18%) gray overall.



Edit: Nikonite @WayneF has provided us with this excellent tutorial on the subject of how light meters work.
....

 

[J-see]

Thanks, I'll have a read and do some testing tomorrow. When shooting clouds I see an immediate difference between center and matrix since I use it to focus on the lightest parts and in that "invert" the light/darkness. Center does a better job at it since, as I (maybe incorrectly) assume, it only measures the center and adjusts all accordingly.

It'll take some to convince myself it isn't overexposing.

 

[WayneF]

The middle grey result is correct. That is a big feat for a meter. Meters don't have any human smart to recognize the scene or the subject or anything. They don't know what it is, or how it ought to be. They just see a blob of light. They will put it in the middle, not too bright, not too dark. then the details are up to us.

But technically, it is not 18%. Nikon, Canon and Sekonic use 12.5% (spec in the Sekonic manuals). Minolta, Pentax, and Kenko (bought Minolta meters) use 14%. That's a very minor difference.

Kodak always used to say (published instruction sheet) to open 1/2 stop more if metering from their 18% gray card, which makes that result be about 12%. Kodak sold off their gray card business nearly 20 years ago, and this little note was lost. 18% was a Ansel Adam thing, about his Zone System in the 1930s.

Reading material:

kodak 18% gray card open 1/2 stop - Google Search

 

[Horoscope Fish]

But technically, it is not 18%. Nikon, Canon and Sekonic use 12.5% (spec in the Sekonic manuals). Minolta, Pentax, and Kenko (bought Minolta meters) use 14%. That's a very minor difference.

Kodak always used to say (published instruction sheet) to open 1/2 stop more if metering from their 18% gray card, which makes that result be about 12%. Kodak sold off their gray card business nearly 20 years ago, and this little note was lost.

Interesting!

....

 

[J-see]

I did the simplest test there is. Pure white, pure black and white next to black. If the cam overexposes all the time, white and black should be different. They're not.

This is how they come out of the cam using auto-WB. The second series are after LR's auto-WB.










Slight difference but not enough. I was wrong.