This is the first exercise for light, and is designed to start us to think about light as a separate element in photography - a controllable input as opposed to a given factor.
We are looking here at exposure specifically. In a way, it is back to the days of light meters and measuring exposure at different points in a shot to see what is optimal. Post processing technology has to a large degree taken over the need to get exposure exactly right on scene, although it saves time to get it right first time, and even the most advanced programs cannot deal with a shot that is a long way from the ideal exposure - noise rapidly gets introduced.
For the first part of the exercise, I took two images deliberately under exposed, and three over exposed. The first two were taken near my home:
 |
| f4; 1/4; EV-2 |
Although taken late in the day, I reduced the EV in order to convey the impression of dark, slightly spooky wood leading to the light.
 |
| f4; 1/60; EV-2 |
This image of a tunnel appears darker on the blog than on screen in Adobe Bridge (one further factor beyond the control of the photographer is the variable brightness of monitors, projectors and software). Again, I wanted to emphasise the slightly threatening environs, in particular keeping the tunnel end virtually black so, unlike the first image, the exit looks to offer no relief.
Three brighter images were taken on the Bristol-Bath railway path:
 |
| f4; 1/125; EV+1 |
 |
| f4; 1/125; EV+1 |
The first two images are of sculptures inserted in the window frames of the only surviving wall of a former station on the line. It was important to increase the exposure so as to bring out the detail and colour in the sculptures, which are each partially shaded and back lit.
 |
| f4; 1/6; EV+1 |
The camera was placed on a bag for this image of the Staple Hill tunnel to avoid camera shake on a slow shutter speed. The cycle path runs bottom right to exit mid left. I thought that an overexposed image would convey the opposite message to the tunnel image above: that it is well lit and quite nonthreatening. Frankly, it is overdone - see the same image with about 60 seconds' worth of adjustment in the Adobe RAW processing software: the image is converted to black and white, exposure is reduced, contrast boosted and the tone curve tweaked; an altogether better image.
Next, we are asked to take five or six photographs, each with five exposures arranged around what is the best exposure. I had no tripod for any of these (although the tunnel images were taken by placing the camera on a rock) so the alignment is not precise. I refrained from using the auto align function in CS5 as geometric precision is not the point of this exercise. Constant aperture was maintained in order to achieve same depth of field; the shutter speed altering in each case. ISO is 200 for every image. I have the camera set to each EV increment to be 1/3 of a stop so, for example, EV+1 is 2/3 of a stop difference (as the stop increments are in halves). In each set of images, the exposures are successively brighter in the following sequence: EV-2, EV-1, EV0 (judged to be the optimum), EV+1, EV+2.
I took three sets of images on the Railway path:
 |
| f4; 1/500 |
 |
| f4; 1/250 |
 |
| f4; 1/125 |
 |
| f4; 1/60 |
 |
| f4; 1/30 |
The brighter images do not work well - EV+2 is washed out and EV+1 simply too bright. EV0 works just fine here, the underexposed images being too dark around the hedges. There is little sky so the fact that it is overexposed has little impact.
 |
| f4; 1/0.7 |
 |
| f4; 1/0.7 |
 |
| f4;2sec |
 |
| f4;4sec |
 |
| f4; 8sec |
As with the image above, the overexposed images do not work - look artificial and texture is lost. There seems to be little difference in the darker images, although unquestionably the EXIF data says one was taken at EV-2 and the other at EV-1. Possibly the EV-1 is preferred - it seems more natural.
 |
| f4; 1/1000 |
 |
| f4; 1/500 |
 |
| f4; 1/180 |
 |
| f4; 1/125 |
 |
| f4; 1/60 |
Again, the brighter images of this sculpture are washed out - detail and texture lost. Even the EV0 image looks brighter than perhaps is optimal. My choice would be the EV-1 image, although the colours (remember how the impact of brightness on colour was discussed before?) are most natural in the EV-2 image.
I took the opportunity of an alumni weekend at Cambridge to get some shots of the colleges for the other set of three:
 |
| f13; 1/500 |
 |
| f13; 1/250 |
 |
| f13; 1/125 |
 |
| f13; 1/60 |
 |
| f13; 1/30 |
This shot is of St Catharine's college quad. It is an ideal candidate for HDR shooting as the exposure is perfect of the shaded areas in the EV+2 image, whilst the sky is optimized in the EV-2 image and the college walls in the EV-1 image. Average metering cannot cope with this dynamic range. Overall, the EV-1 image looks best.
A similar issue with this classic English shot of King's College chapel with some foreground interest of the sheep. The two extremes are not good. The EV+1 image does justice to the shaded foreground without burning out the building; albeit the sky is burnt out. EV-1 image has a pleasant sky and a little more colour depth consequent on the darkening of the facing wall and the foreground is just visible.
The final image is of Clare college, taken as an organised run was in progress using the college as part of the route.
This is the one image where the brighter EV+1 shot probably is the best. Helped by the more interesting foreground, the shadow is lifted nicely; sky is too bright but retains some colour. The shadow in the two darker images is too strong, dominating the eye movement. The sky is pleasant in EV0 image, but shadow is still too strong in my view.
Quite a big exercise, this reinforced my preconceptions that using brighter exposures rarely works for my camera. Even for inside shots, increasing exposure tends to slow shutter speeds well below one that can be used with hand held camera. I usually shoot outdoors with EV-0.5 or EV-1, a policy supported by these images. The main point reinforced to me was the impact of brightness on colour and tonality. It just adds to the conclusion that photography is about the whole image and balancing a number of constituent elements.
