Exposure Part II - Highlights

In part II of my guide to exposure I look at how to handle highlights and shadows.
In Part I of my short guide to exposure, I explained how to fix a good midpoint using a spot meter. Following this basic technique will get you, consistently, an image with well exposed mid tones.

But what do you do about situations without a clear, useable mid tone? For example, how do you meter for a black rock on snow? How do you get the snow to be white? In most cases an automatic meter will turn the snow grey. Or it will blow the highlights so that the snow is white with no detail. How do you get the snow to be white with detail, every time? The answer is you need to know how to work the shadows and highlights of your image.

Let’s take the snow example. First, meter on the brightest spot in the snow. The meter will assume this is a mid grey so you will need to increase the exposure so that it is on the extreme edge of the latitude of your film/sensor – in other words just before it goes completely white. But by how much do you increase exposure?

To know this you need to know the dynamic range of the film/sensor you are using. In other words how many stops of light there are from the black point to the white point.

For example using Velvia colour transparency film, there are 4 stops between black and white. The white point is 1.7 stops above the mid point. So if you want to get details in something that is white (e.g. a white building or snow) meter from the highlight and increase the exposure by 1.7 stops. For example 1/60 at f11 becomes 1/20 at f11. Or 1 sec at f22 becomes 1 sec at f11 1/3.

If you are using a digital sensor you need to increase the exposure by 2 (possibly 2.3) stops. Meter on the snow and add 2 stops so that for example 1/250 at f11 becomes 1/60 at f11. Bingo – white snow.

I use this approach to ensure that the highlights in white flowing water are not blown – very easy to do. I will spot meter on the brightest area of white water and add +1.7 for film and +2 for digital. For example, see my image 'Rowan Tree Falls'.

Similarly if you want to guarantee shadow detail on film, meter from the shadow and reduce the exposure by 2.3 for velvia. So for example, 1/60 at f11 becomes 1/320 at f11. On digital reduce the exposure by 3 stops.

To summarise, to meter for the highlights here is what I use:

  • Velvia +1.7
  • Provia +1.7
  • Digital +2
For shadows I use:
  • Velvia -2.3
  • Provia -2.7
  • Digital -3

You can practice this with white paint on buildings in sunlight using a DSLR. Again, quite quickly you will get accurate results consistently and it will become second nature.

You should use this technique to check where the brightest point of your image will fall when you have fixed a midpoint. I do this every time. To do this find the brightest point (often the sky or a water reflection). On transparency film like velvia, if it falls >1.7 stops from your intended midpoint exposure, the highlights will be blown. For example, if the mid point is 1/125 at f11 and the highlight is 1/125 at f32, there are three stops of light between them and the highlights will be pure white. No nice fluffy clouds.

This is where a handheld spot meter comes into its own – I can spot the midpoint and then spot the highlight and visibly check at a glance whether it is inside the dynamic range of the film. It takes me seconds. Similarly on digital I can check the highlights in seconds – I take a test picture and check the histogram to see if the highlights have blown.

Quite often a good midpoint will mean blown highlights or no shadow detail, especially in landscape photography. Why? Because in many situations the brightness range of a scene (particularly one that includes the sky) is much more than the film or sensor can cope with. (e.g. 5, 6 or even 7 stops of light or more.)

What do you do in these situations? Well you can let the highlights blow but in practice this makes for ugly images – great patches of featureless white. What is more, on digital, you can’t recover lost highlight detail later in post processing like you can the shadow details.

The first sensible option is to meter for the highlights, rather than a midpoint. Find and meter the brightest spot. If you want it to be white with detail, increase the exposure by +1.7 for film. Use a bit less if you want to retain the highlight's colour (e.g. for a grey cloud I use +1).

This will protect the highlights but at a cost – the image will end up overall darker and shadows might block up on film. Here you use your experience of what the resulting image will look like to judge if it will work. (I would check the shadows and see if they are in range). On digital I might judge that I can fix it easily, afterwards, in post processing using curves.

The second option is to switch to a higher latitude film like colour negative – it has 8 stops of dynamic range. I always carry a few sheets of colour negative film – it is very useful and sometimes it is nice to have a lower contrast image.

Thirdly, there are various digital techniques to increase the dynamic range involving blending multiple images. The approach works but it has its limitations. A web search on HDR or image blending should find plenty of tutorials.

I use the first two options regularly and the third occasionally. However for the serious landscape photographer there is one further option – graduated filters. These are used to darken parts of the image to bring the highlights into the dynamic range of the film or sensor. This allows you to meter based on a good midpoint and still have highlight detail.

How do you use them? That is the subject of Part III.

© Jon Brock 2007