Getting it right the first time




One of the most complicated technical issues in photography is getting your exposure right, in other words allowing just the right amount of light to reach your camera’s sensor. 

"But is that not mastered by the camera itself? Just use Auto exposure, Auto ISO and the build-in Multi Field Matrix whatever it is called metering mode".

That will work indeed, but then you are giving away all control of one of the most important creative and technical aspects of the art and craft of photography.


As an example, imagine a kitchen with a very large window and two persons having a conversation in front of the window and you want to take a series of photographs. All the light is coming from the window behind the two persons and this light will not change during the session.


Not Identical

When using Auto exposure most of your photographs will receive a slightly different exposure due to the fact that in some photographs the persons will be covering a different part of the window (if at all), depending on your shooting angle, composition etc. The camera will interpret the scene for every shot in a slightly different way due to the fact that the ‘matrix’ is interpreting the scene as different objects for every shot. When you throw in some headshots the differences will be even more pronounced.

All this does not make sense. The scene did not change, the subject did not change, so the exposure should not change either. Even worse, in digital post-production you would have a lot of work to make sure all photographs end up having an identical look. When you only shot JPEG you would be in real trouble, all photographs would look a little different and could not be used together as a series.


What is ‘exposure’

We are actually talking about measuring the light and translating that measurement into the right amount of light reaching the camera sensor.
Most modern camera’s have a built-in light meter with the option to select different ways to measure the light, usually a matrix mode, a centre-weighted mode and maybe a spot metering mode.

Good, one problem solved, we have a light meter.
Yes, but unfortunately one that requires a lot of knowledge from the user to be used correctly.
Let me explain.

Any light meter will measure whatever it is aimed at and will translate the measured amount of light into a result that will look like “18% grey”. Please check out the link to understand what 18% grey means and why it makes sense to use that standard.

This will work for the average subject, but imagine measuring a black velvet cloth. Your light meter wants to translate the dark black into grey and as a result will give you a measurement result that would over-expose the photograph. The same story for a white piece of paper, the meter will want to transfer it into grey, resulting in an underexposed picture.

A good intelligent in-camera measuring system like matrix metering will compensate, but the results are usually far from perfect. If you try this with a centre-weighted or spot metering mode you will observe the above described results.

Conclusion: Any light measurement needs to be interpreted, preferably by an experienced human being.


Dynamic range

And that is only the first part of the problem. Now it gets a little complicated.

Your sensor can handle a certain difference between dark parts and light parts in a subject. 
If the subject reflects or submits more light then the sensor can handle the result will be a light part without any detail. 
If the subjects reflects or submits less light then the sensor can handle this dark part will have no detail whatsoever. We call this range between detailed blacks and detailed whites the dynamic range of a sensor. 
This is a natural given property of the sensor design.

Dynamic range has nothing to do with what can be recovered from a RAW file in post-production other then that you cannot recover what is not there, so everything outside the sensor’s dynamic range will be lost forever. Recovering a RAW file just means getting the full range out of the sensor.

Your subject will also have a dynamic range, called the luminance range. Imagine the kitchen scene described earlier. The world outside the window will reflect a certain amount of light and the shadows inside will reflect a totally different, much smaller amount of light. The difference between the biggest and the smallest reflection of light is the luminance range.


Exposure Value

Now I need to introduce another phenomenon in photography, the Exposure Value or Stop, and therefore we have to discuss aperture, shutter speed and ISO.

When you make a photograph you always use the aperture (in the lens) and the shutter (usually in the camera). 
You could imagine the aperture as an opening that can change it's diameter and the shutter as a time device that allows the light to reach the sensor for a long or short amount of time.

When the camera is set to fully automatic it will set the aperture and shutter speed for you, otherwise you are allowed to select one or both yourself.

Example

A specific combination of shutter speed and aperture will result in a specific amount of light passing through the aperture and shutter to the sensor. Let’s call that value “10” for now, just a number, no photographic significance yet.

If you open up the aperture one click you double the amount of light passing through. The value of light reaching the sensor is now “20”. 
When you close the aperture one click, as it was before you changed anything, the value of light passing through becomes “10” again.

You can do the same with the shutter speed. Change the shutter speed to half the value of what is set. You have now doubled the amount of light passing through the shutter, changing the value to “20”.

To reach a value of “5”, what would you have to do?
- Close the aperture
or 
- Set the shutter speed to a faster value

And what would happen if you close the aperture one click and set the shutter speed to half the value? 
The aperture goes from light value “10” to “5” (you closed it)
The shutter speed goes from light value “10” to “20” (you allow the light to work it’s way onto the sensor for the double amount of time). 

The netto result is a light value of “10” to reach the sensor, in other words the total amount of light remains the same.

ISO

The ISO value represents the sensitivity of the sensor. 
It is a set value, changing the ISO value in your camera will only trigger an electronic calculation that will alter (usually boost) the electronic signal coming from the sensor. It is basically the same technique as amplifying a microphone signal to fill a stadium for a rock concert.

EV

This number “10” I used for the examples does sort of exist in photographic technique, it is called the “Exposure Value” and any decent light meter will allow you to calculate it.

For instance, with an ISO of 100, an aperture of f4.0 and a shutter speed of 1/60 of a second the Exposure Value (or EV) is 10.
With the same ISO when setting an aperture of f2.8 and a shutter speed of 1/125 the EV remains the same,10, because the amount of light passing thru remains the same.

The only difference with the above example is that doubling the amount of light means changing the EV from 10 to 11, not to “20”.

Each difference in EV is called a ‘Stop’. Going from EV 10 to 11 is the same as one ‘Stop’.


Measuring the light

Back to our subject and the luminance range.

Let’s assume that the sensor of your camera can cover a dynamic range of 10 Stops.
The subject of the kitchen scene will probably have a luminance range of more then 10 stops, remember the world outside the window and the shadows inside. Let’s assume a luminance range of 20 stops.

Now we have a problem. Your sensor can never register the complete scene, it will always lose some detail somewhere, it will somehow loose 10 stops.

A choice will have to be made. What part of the scene do we want to loose, the shadows, the world outside the window, a little of both? Remember, we can’t have it all.

Matrix

Your matrix meter will make this choice for you. 
It will measure all lot of squares in the subject (the matrix) using the 18% grey method and will then compare the measured total result to a huge database of existing values in it’s memory. These values represent existing photographs translated into matrix values and the camera will select the best matching combination and expose accordingly. 

It is very wel possible that the camera actually thinks it is looking at a sun-lit snow scene instead of a kitchen with a window. 
When you change your composition the value of these measured matrixes changes and the camera will re-calculate the scene, maybe thinking it is now looking at something like a summer scene outdoors with people in it. 

Therefore Matrix metering can not be trusted, it will re-calculate for every photograph. It may calculate correctly, it may not. You have no idea what is measured and how that is translated into an exposure. What part of the luminance range is dropped? You don't know.

Center-weighted

You can set your light meter to the Center-weighted mode. 
Now all intelligence is switched off. The camera measures the center of what is seen by the lens, usually about 60 percent of the total picture. This will result in a reproducible measurement with a predictable result. Very good. 

Only problem, if you have no experience the end result will probably be a photograph that is not as correctly exposed as the one with the Matrix metering.

Spot

When you use spot metering you set the camera to measure a very small part of the scene. This will allow you to determine the exact luminance range of the subject and make a very calculated decision about the correct exposure. 
Unfortunately, if you are not very experienced this will usually result in very badly exposed photographs, it is a complicated technique.


Now what

Oh dear.
What to do? All metering modes seem to have an issue.
Indeed they do.

Choose

Let’s get back to the luminance range of the subject and the dynamic range of the sensor. They did not match.
But do you really want to catch the complete scene in your photograph? Is it that important that you show the outside world as seen through the window + have all the details in the shadows? You are photographing two persons who happen to be in a kitchen. The persons are the main subject.

Solution

A possible solution could be to set the camera to Center-weighted, take a step forward and measure the two persons only. 
Take a step back so you have the original scene again, set the camera to manual, set the aperture and shutter speed according to your measurement and make one test exposure.

Check the result on the screen of your camera, the histogram may give you an indication. Do you like the result? If not, change the exposure to your liking (persons to dark = allow more light to enter the sensor etc.).

Take another test shot. When you like the result keep the exposure settings like this for the entire session. When the subject changes, i.e. because the persons moved into the garden, measure again, test again, set the new exposure and change nothing during this new scene.

Use your screen and histogram to check the exposure after measuring the light. Don't check the photographs you just made, don't chimp, you don't want to miss good photographs, concentrate on the subject, not the camera.

Keep the EV

This is the most important part.

Measure the exposure using whatever method that works for you, set the EV value manually on the camera using the ISO, aperture and shutter speed and don’t change these as long as the scene / subject / light does not change. 

You can only change aperture, shutter speed and ISO as long as the resulting EV remains the same (open one, close the other etc.)

No auto-ISO, no auto aperture or auto shutter speed or program mode. MANUAL.
Only change a setting when the light changes.

This is very simple and with some experience becomes second nature. Never use Auto anything. You want to be in charge.


Different ways to measure the light

We have seen that the way the camera measures translated everything into 18% grey.

We call this method ‘reflected light measurement’.
The light meter measures the light that is reflected by the subject.

All versions of reflected light measurement suffer the same issue, is the measured subject 18% grey, and if not, how do we interpet the measurement?
There are some solutions though.

Grey card

One option would be to measure not the subject, but a special grey peace of plastic or cardboard that is indeed “18% grey”, you can buy these. 
Now we do measure 18% grey so we know exactly what our measurement is worth.

Warning: There are also grey cards that are meant to measure the white-balance of a scene (another useless concept when shooting RAW). These are NOT 18% grey and cannot be used for this light measuring method. You need a special 18% grey card.

"Light meter"

Another option is to use a light meter that does not measure the subject but measures the light falling on the subject. This is a specialised technique but the result is roughly comparable to the result of measuring a 18% grey card.

"Light metering" does not measure the subjects luminance range and that could be a problem.

Spot metering

This is a very specialised type of reflected light metering, but it is the most reliable when executed correctly.

When using spot metering, we measure the luminance range of the subject and / or specific parts of the subject.

You usually point the ‘spot’ at the darkest part of your subject, note te value and do the same with the brightest part of the subject. You now know the luminance range.

Next you take another measurement of a very important part of the subject, like a face or a grey card if available. 

You can now make a very educated guess about the correct exposure

All this takes some time to execute and you will need a lot of experience and trial and error to make this work, not very suitable for anything other then some type of studio / controlled setup. You can buy specialised light meters that allow spot metering as part of a studio setup.


Real life

How do we handle all this in real life, photographing our dynamic and ever changing subjects?

Measure light using Centre-weighted metering once, set the ISO, aperture and shutter speed manual and leave it as-is. Re-measure from time to time, change when light changes. 
This is actually a lot faster then auto-expose > chimp the screen for the result > compensate exposure using exposure compensation > do the same for the next photograph > repeat again and again.


Some tips

No change

Light does not change that often.
When the light does not change, your exposure does not change. 

It is very wel possible that you measure the light once when starting your photography day / session and not change anything during the rest of the day / session, because the light remains the same. No headache about exposure anymore.

Skin

Caucasian skin is usually one stop lighter then 18% grey.
Measure the inside of your hand, compensate one stop, done. 
(subject to light = underexposure = allow more light then measured)
Test this once at home for your own skin and you are set.
You could even spot measure the subjects skin and compensate one stop. Very fast and quite accurate, but please do some testing first.


Exposure compensation

What about exposure compensation and your camera’s option to freeze the automatic exposure between shots?

Automatic exposure was invented for people who do not know how to measure light. Photography used to be difficult, ‘did the picture come out right’, in other words did you expose correctly? Most people did not and the results would be variable to say the least. Auto exposure changed all that and it is indeed a blessing for the uninitiated. It does not work if you want to control what you photograph. Matrix metering is a (big) improvement on Automatic exposure, but with its own drawbacks.

Exposure compensation and freeze exposure are just ways to give some sort of control to the user, but as we saw that matrix metering is flawed it does not make sense to correct an exposure that we don’t understand / control anyway. 

A lot of options are added to cameras to compensate for flawed techniques or uneducated users, auto exposure & exposure compensation are part of that, autofocus is another. So called youtube reviewers that miser about exposure compensation etc. know nothing.

Guesstimate 

In the old days photographers did not measure any light. 
They would guesstimate the right exposure by looking at a scene. 

That is actually a nice exercise, just make a guess about the right exposure and then measure the light. Your results will improve rapidly when you repeat this exercise every time you measure light, the technique becomes quiet reliable with practise.

If you look at pictures of older photo journalists at work you will notice that they used camera’s without any light meter (Cartier-Bresson, Alfred Eisenstaedt, Robert Capa, Margaret Bourke White to name a few)
They all used this estimate technique and made great photographs, it really works.

Measuring light became normal practice around the Vietnam war era. Don McCullin says he did not want to risk his life for an incorrect exposed photo so he always measured light.

Journalists

Matrix metering (and Auto-focus) on professional camera’s is there for the photo journalists. They sometimes have no time to measure anything and just need a photograph that is ‘as perfect as possible under the circumstances’.



Photograph

The photograph on top shows a camera with the perfect centre-weighted light meter, a Leica M6 + a real 'light' meter, recognisable by the white sphere. This combination will allow you to measure the light fast and correct every time. When used in the right way it is much faster then any modern DSLR or Mirrorless camera.



More

You might want to read 'inside raw files' on digitalphotopro.com for an in-depth explanation on RAW files

Comments