Measuring exposure - part 1

In August we spent a week in Edinburgh for the first week of the Fringe Festival, followed by a few days near Peebles and five days on a walking tour along the central part of Hadrians Wall. In addition to taking the opportunity to visit three photography exhibitions taking place in Edinburgh, I was also taking photographs for the first exercises of the Light section of this course.

Having thought about the instructions for the measuring exposure exercise, I believed that I should look for situations where a deliberate adjustment of exposure was necessary to obtain the intended exposure.

I have cropped, cleaned up dust spots and sharpened these photographs, but I have not adjusted the original exposure or contrast settings. However, in two cases, I have made adjustments to copies of the originals where I wanted to make a comparison.

Image 1: View of the Pennines from Hadrians Wall Path

Nikon 18-70mm lens at 25mm: Iso 200 f8 at 1/750th second (Exposure minus 1 stop)

This is a very high contrast scene with bright clouds and sky, but dark patches of trees and distant mountains, especially in areas shaded by clouds. I took a sequence of three photographs of this view, the first with zero exposure adjustment, the second -1/2 stop and this one -1 stop. There was originally a field and trees in the foreground filling the bottom third of the frame.

In the first two photographs, the field in the foreground was mid-tone, but the clouds were severely blown out and looked unreal. In the third photograph with a -1 stop adjustment, the foreground field and trees were too dark, but the clouds looked just right. I felt that the clouds and sky were significantly more interesting than the foreground field and trees, and I therefore cropped away almost all the dark area of trees and fields in the foreground leaving just thin line along the bottom of the frame.

I generally keep my camera set to "pattern" which is the equivalent of matrix type exposure metering. In this case, I assume that the camera's meter was giving more weight to a mid-tone exposure of the land and less weight to the sky. As a result, the bright sky was severely overexposed until I adjusted the exposure to hold the detail in the clouds.

Image 2: Neidpath Castle, Peebles

Nikon 18-70mm at 22mm: Iso 640 f8 for 1/90th second (Exposure minus 1.5 stops)

I saw this view one evening from the road into Peebles as the sun was still catching the top section of the castle. This is another high contrast scene with a dark conifer and shady foreground, but with bright sunlight falling on the castle and trees in the background.

Without thinking too carefully, my intention was to take a bracketed sequence of photos from -1 up to +1 stops exposure. However, the camera screen and histogram showed me that at these settings the castle was totally overexposed. In order to keep some detail in the stone walls of the castle, I needed to reduce the exposure to - 1.5 and -2 stops. Even at these settings the sky was totally white and featureless and I have cropped out almost all the sky in the above picture.

As usual, the camera was set to "matrix" metering and I suspect that the exposure was mainly influenced by the very dark foreground trees, rather than the brighter middle ground and bright sky, which originally filled the top quarter of the frame. I wanted to keep the dark tree in the foreground as this conveyed the thick forest around the castle which hides it from view.

Image 3: A section of Forth Bridge

Nikon 18-70mm at 46mm: Iso 125 f7.1 for 1/400th second (No exposure adjustment)

The photograph was taken from the harbour at South Queensferry between the two Forth Bridges. There was virtually no wind or waves which meant that the reflections of the pillars of the railway bridge were unusually clear. This is a high key scene with the light coloured sky above reflected in the water in the foreground. As a result, the camera's exposure measurement has underexposed the photograph making the colours more subdued than they should be.

I should have anticipated this and increased the exposure with half stop or full stop adjustment before taking the photo.

To gauge the effect, I have included a copy of the original photograph in which I have increased the exposure by one stop:

Image 3 copy with 1 stop increase in exposure

After adjustment, the brightness of the day is more natural.

Image 4: South Queensferry harbour at low tide

Nikon 18-70mm at 70mm: Iso 125 f7.1 for 1/350th second (no exposure adjustment)

This scene has a dark foreground and mid-ground, with a band of lighter colours towards the top of the frame. Because of these predominant dark colours, I thought that the matrix exposure meter would tend to over-expose the photograph. I therefore expected that I would need to make a negative exposure adjustment. However, the first photograph with zero exposure adjustment already looked quite dark and the histogram showed that shadows were starting to loose detail. I did not therefore make any further adjustment.

Subsequently examining the photograph in Lightroom, I can see that the shadows under the boats have lost detail, whereas the white decking has only small spots of overexposure. This suggested that the photograph would have been improved if I had increased the exposure. Again, I have prepared a copy of the photograph in which I have increased the exposure by one stop. Whilst more of the areas of white decking have become totally white, these are flat plain surfaces without visual texture and I do not feel that anything of importance has been lost. The mud however is brighter and more detail is revealed in the shadows under the boats. On balance, I prefer the photograph with increased exposure.

Image 4 with 1 stop increase in exposure

In this case, I have been unable to work out why the matrix metering chose such a low exposure for a relatively dark scene.

Image 5 North Bridge Edinburgh

Nikon 18-70mm at 18mm: Iso 400 f6.3 at 1/10th (exposure minus 1 stop)

I took this photograph at twilight, pointing away from the setting sun. The windows of some of the buildings are reflecting the brightness of the sky where the sun had set, enlivening their otherwise rather somber appearance.

I had increased the Iso to 400 as the light levels were low. I had set the exposure to minus 1 stop for the previous photograph and had forgotten to change it. I expected the image to be blurred as the shutter speed was too slow for a hand held shot. However, it turned out that the photograph is not too blurred.

The tarmac and buildings are relatively dark in tone, with a smaller area of sky. The matrix metering would therefore be likely to make the tarmac and buildings too light. The reduction of 1 stop records an appropriate level of dusky gloominess so that the small highlights of the windows and street and vehicle lights stand out.

Conclusion

Whilst I was aware of some of the situations where the camera's meter can misread a scene from my days of using film, I often overlook checking this before pressing the shutter with my digital camera. It has been useful to revisit and understand better this process.

I have started to check the histogram on the camera screen, and I am surprised how often I find something unexpected.

Another point that I learned was the power of image processing software to selectively adjust lights, darks etc. In order to create adjusted versions of the photographs that I used for comparison in my notes, I used the "exposure" slider in Lightroom which applies a universal adjustment to the whole picture. Normally, I use the more selective sliders such as "brightness", "fill" to produce the final version of my photographs. Had I used these, it is relatively simple to restore detail to both shadows and highlights in the harbour photograph for example.