Metering

[tdodd]

A meter, of whatever kind, makes a guess. If you know what you're doing, usually a good guess, but it's easily fooled.

Ignoring evaluative/matrix metering for now....

Meters don't make guesses. They make measurements. To say the meter makes a guess is as daft as saying that a thermometer makes guesses, or that a car speedometer does, or a clock.

A thermometer does not tell you if it is too hot or too cold. It tells you the temperature. You get to decide whether that's right or wrong.

A speedo does not tell you whether you are going too fast or too slow. It tells you the speed you are doing. It's your decision whether the speed is appropriate.

A clock does not tell you whether you will be early or late. It tells you the time now. Is it past bedtime? Not a decision for the clock. Will you be late for the meeting? Not the clock's responsibility to decide. You read what the clock says. You decide how to interpret the data it provides. Is it lunchtime? Possibly. Are you hungry? How the frack should the clock know?

A camera meter does not make guesses. Stick the camera in manual exposure mode, meter (part of) the scene and the meter will tell you how much brighter, darker or equivalent to "middle grey" the part of the scene you are metering will be recorded. You get to choose how much brighter, darker or equal to "middle grey" you want that part of the scene to be.

If I spot meter my own palm with the camera at +1 1/3 is there guesswork involved? If I spot meter highlights at +3 is there guesswork involved? If I spot meter green grass at -2/3 is there guesswork involved? If I decide that there is no bright content (of importance) within my scene then I may increase the exposure by 1 stop in order to capture more shadow detail. Is that guesswork by the camera or by me? No. There is measurement, conscious decision making and total control over the exposure. IF there is any guesswork then it is by me, not the meter.

Does a Sekonic incident meter make guesses? What about the flash meter? What about the spot meter? None of them make guesses; they measure.

Now, if you're going to leave the exposure setting entirely up to the camera, without providing any input yourself, then sure, you could make the case that the camera is "guessing" at the right exposure, but really it isn't, except in the case of evaluative/matrix metering. In all the other metering modes it is quite dumb, assumes nothing, guesses nothing and is simply resolving the metered area to a "middle grey" tone. That's not guessing. That's maths and science.

There is a missing ingredient from this and that is human involvement, providing instruction to the camera on how much brighter or darker the metered element is than "middle grey". That's why cameras provide an exposure compensation control for autoexposure modes and a +/-3 stop (+/-2 stops on some cameras) meter to support manual metering. For anyone to suggest (and many do) that a correct exposure is indicated by a centred meter needle is as often as not a complete nonsense. It would be true if the metered area was equivalent in tone to "middle grey". For anything else it would be wrong advice. Sadly I've seen that advice spouted far too often.

Evaluative/matrix metering does result in a more considered outcome, but then the camera is making a judgement (guess) about what sort of scene you have. Quite frankly, based on the need to dial in EC for things like BIF and aircraft it isn't as intelligent (not as good at guessing) as photographers might wish. That's why I prefer to let the camera take care of scientific measurements (using spot metering) and I'll supply the intelligence (guesswork if you like) required to achieve an optimum exposure. The meter is guessing at nothing. People who get poor exposure results from their equipment either have faulty equipment (probably not likely) or are relying too much on their gear and not enough on themselves.

Chimping histograms and blinkies is all very well, and I make use of those features too. But it's a bit late to do that once the decisive moment has passed, which is why knowing how to meter before taking the shot is a skill worth having. Fortunately for me I usually prepare in advance, which means I not only meter and set exposure beforehand, but, if I have doubts, I can often fire off a test shot just to check.

I may have said it earlier (I'm not re-reading the whole thread) but if you want to chimp your results it is IMHO a waste of time to evaluate your exposure from the image on the back of the camera. Blinkies are a useful indicator and so is the histogram. The rest of the image is not. Use the image for assessing composition and possibly focus, not exposure, not white balance. It will probably mislead you.

FWIW I am always trying to improve my own metering skills, because I certainly don't get it right first time every time. The better I become the more pleased I am with my skills as a photographer. Chimping an image until it is correct gives me no sense of pride at all. There is no skill involved there. Any fool can chimp an image until finally it is correct. Nailing a shot first time (every time) - that's photography. Maybe one day I'll get to call myself a "Photographer". Right now I'm still practicing to become one.

 

[pampers]

Blimey.
What a responce. When I made the post, I had no idea that so many people would, take the time and effort to give so much good and detailed advice.

I have taken as much as I can on board. And hopefully will be able to get out and try, some of the things that you all said.
And if needed I can always re-read it.

A realy BiiiiiiiiiiiiiiG thank you to everyone

 

[HoppyUK]

Ignoring evaluative/matrix metering for now....

Meters don't make guesses. They make measurements. To say the meter makes a guess is as daft as saying that a thermometer makes guesses, or that a car speedometer does, or a clock.

A thermometer does not tell you if it is too hot or too cold. It tells you the temperature. You get to decide whether that's right or wrong.

A speedo does not tell you whether you are going too fast or too slow. It tells you the speed you are doing. It's your decision whether the speed is appropriate.

A clock does not tell you whether you will be early or late. It tells you the time now. Is it past bedtime? Not a decision for the clock. Will you be late for the meeting? Not the clock's responsibility to decide. You read what the clock says. You decide how to interpret the data it provides. Is it lunchtime? Possibly. Are you hungry? How the frack should the clock know?

A camera meter does not make guesses. Stick the camera in manual exposure mode, meter (part of) the scene and the meter will tell you how much brighter, darker or equivalent to "middle grey" the part of the scene you are metering will be recorded. You get to choose how much brighter, darker or equal to "middle grey" you want that part of the scene to be.

If I spot meter my own palm with the camera at +1 1/3 is there guesswork involved? If I spot meter highlights at +3 is there guesswork involved? If I spot meter green grass at -2/3 is there guesswork involved? If I decide that there is no bright content (of importance) within my scene then I may increase the exposure by 1 stop in order to capture more shadow detail. Is that guesswork by the camera or by me? No. There is measurement, conscious decision making and total control over the exposure. IF there is any guesswork then it is by me, not the meter.

Does a Sekonic incident meter make guesses? What about the flash meter? What about the spot meter? None of them make guesses; they measure.

Now, if you're going to leave the exposure setting entirely up to the camera, without providing any input yourself, then sure, you could make the case that the camera is "guessing" at the right exposure, but really it isn't, except in the case of evaluative/matrix metering. In all the other metering modes it is quite dumb, assumes nothing, guesses nothing and is simply resolving the metered area to a "middle grey" tone. That's not guessing. That's maths and science.

There is a missing ingredient from this and that is human involvement, providing instruction to the camera on how much brighter or darker the metered element is than "middle grey". That's why cameras provide an exposure compensation control for autoexposure modes and a +/-3 stop (+/-2 stops on some cameras) meter to support manual metering. For anyone to suggest (and many do) that a correct exposure is indicated by a centred meter needle is as often as not a complete nonsense. It would be true if the metered area was equivalent in tone to "middle grey". For anything else it would be wrong advice. Sadly I've seen that advice spouted far too often.

Evaluative/matrix metering does result in a more considered outcome, but then the camera is making a judgement (guess) about what sort of scene you have. Quite frankly, based on the need to dial in EC for things like BIF and aircraft it isn't as intelligent (not as good at guessing) as photographers might wish. That's why I prefer to let the camera take care of scientific measurements (using spot metering) and I'll supply the intelligence (guesswork if you like) required to achieve an optimum exposure. The meter is guessing at nothing. People who get poor exposure results from their equipment either have faulty equipment (probably not likely) or are relying too much on their gear and not enough on themselves.

Chimping histograms and blinkies is all very well, and I make use of those features too. But it's a bit late to do that once the decisive moment has passed, which is why knowing how to meter before taking the shot is a skill worth having. Fortunately for me I usually prepare in advance, which means I not only meter and set exposure beforehand, but, if I have doubts, I can often fire off a test shot just to check.

I may have said it earlier (I'm not re-reading the whole thread) but if you want to chimp your results it is IMHO a waste of time to evaluate your exposure from the image on the back of the camera. Blinkies are a useful indicator and so is the histogram. The rest of the image is not. Use the image for assessing composition and possibly focus, not exposure, not white balance. It will probably mislead you.

FWIW I am always trying to improve my own metering skills, because I certainly don't get it right first time every time. The better I become the more pleased I am with my skills as a photographer. Chimping an image until it is correct gives me no sense of pride at all. There is no skill involved there. Any fool can chimp an image until finally it is correct. Nailing a shot first time (every time) - that's photography. Maybe one day I'll get to call myself a "Photographer". Right now I'm still practicing to become one.

I accept that bit in bold absolutely. There is great pleasure from using skill and judgement to get things right, but that's not what I'm talking about. I'm talking about getting correct exposure, by whatever means at our disposal.

When I say that a meter makes a 'guess' there are two aspects to that.

First, it takes a measurement as you say, that should not be wrong, but that assumes that the meter is reading correctly (and is properly calibrated to the camera in the case of a hand meter). Exactly accurate readings and precise reciprocity from very low to very bright light is rare. And there is a strong element of guesswork invloved because the unadjusted reading assumes that the scene will average out to that elusive 'mid grey'. But let's leave that to one side, we know about that, and simply call it the human element of judgement, which must always be applied - sometimes a lot.

The other apsect of the guesswork involved are the assumtions that the meter makes when translating the light level it measures into camera settings. There is a whole raft of them.

It assumes that the moveable feast that is the ISO rating is correct. It assumes that the lens aperture will close down accurately every time (it often doesn't, especially at higher f/numbers). It assumes there is no vignetting (which there very often is*). It asumes that the f/number and the T/stop are the same, which they rarely are on a consumer zoom. It assumes that the shutter is running accurately, which to be fair they usually do these days, though my 5D2 is both a bit slow and slightly inconistent at 1/8000sec.

This is not just theoretical or hypothetical. If you use a hand meter, in a variety of situations, you will undoubedly come unstuck as a result of these entirely unpredictable errors that the meter takes for granted. If you're unlucky and they also get compounded, you can be a long way out. The on-board TTL meter takes account of some of these things, but certainly not all of them. On the other hand, the LCD/histogram/blinkies do of course show the net effect of everything and they show it exactly.

BTW, I have also tested a few car speedos in my time and they are neither accurate nor consistent, for many similar reasons. They calculate speed through assumptions made about gearing (usually pretty safe ones) but also tyres. The latter are significantly affected by rolling radius calculations, dependent on tyre size, pressure, dynamic diameter, and tread wear. Manufacturers calibrate them to over-read by 10% or so, to stop you getting caught out. Some do a better job than others but if you put a GPS device in three different random cars and drive them all at 30mph and 100mph, you will get six different numbers.

*Vignetting. Just this morning I took a series of pictures using a Sigma 50mm and at f/1.4 I needed 2/3rds of a stop more than at f/2.8 because only the centre of the frame was actually anywhere near where it should be when wide open. I usually have peripheral illumation correction enabled so I hardly ever see it when chimping. Of course that's not available with third party lenses and I have to say I was a bit shocked at what I was getting. But at least I could see it on the LCD, and correct it. The meter appeared to be blissfully unaware.

 

[garryg163]

Grass is a fantastic "Substitute reading" for a grey card. Tricky backlit, or high amounts of bright or dark areas in the frame will confuse your meter (ie shooting the red arrows at an air display) Using a grass reading will give a great average reading to give you a chance at a usable shot.