Tuesday 11 September 2012

Exercise 3: Histogram


Before starting this exercise I felt I had a good grasp of what a histogram is - a graph which represents the tonal information in an image in levels of 0-255. A 'well exposed' image has a full range of tones (no gaps either side of the histogram) with the majority being in the middle. A histogram bunched to the left suggests under exposure, and to the right over exposure. This however relies on the idea that there is such a thing as 'correct exposure.' In 'Perfect Exposure', Michael Freeman observes "the best we can say about correctness in exposure is that there are norms that are accepted and expected by most people."1 In 'Photographers Eye' however he points out that "shadows and highlights can contribute strongly to the mood and atmosphere of a  photograph."2 He gives examples of low and high key images that seemingly break the rules of 'correct' exposure and yet are powerful and atmospheric.

For the exercise I had to take images of a low, average and high tonal contrast subject. My first consideration was how I can do this to produce comparable results - I would need to produce a set up that I could manipulate to demonstrate the different conditions required. The exercise also calls for exposures to be taken at +/- 1 EV to compare.

As I was daydreaming one day staring into the distance in my living room I was struck by the way the daylight dropped off against the corner of my window. I was struck with the idea that this could be the subject I was looking for - the corner under daylight conditions I imagined would approximate average contrast and I could introduce artificial light to create both higher and lower contrast.

1: Average contrast

Window light and shadow

0 EV



+1 EV



-1 EV



2: Low Contrast

Light shone on dark side of wall to balance tones

0 EV



+1 EV



-1 EV




3: High Contrast

Strong light shone on dark side

0EV



+1EV



-1 EV



Observations:

The histograms in each sequence behave in the way I expected with the high contrast set showing peaks at either end of the histogram, the average contrast being closer together and the low contrast being bunched in the centre. The low contrast image shows little difference in either the + or - EV shots while the average and high contrast pictures behave in a similar way with the +1 EV shot having less contrast and the -1 EV sot showing more contrast.

The difference between the high and average histograms is less than I expected and when I compare the images the high contrast shots seem much more evidently so than is suggested by the histogram. This is perhaps due to the intensity of the light source used and also the fact that I relied on the camera's matrix metering mode - perhaps I should have metered for the average shot and then set the exposure manually after that. I also had the camera set to automatic white balance and maybe should have also set this to a standard setting. The average contrast appears much warmer than the high contrast image which could partly account for how they are perceived differently.

Another explanation perhaps is that my perception of how much contrast there was in each set up was not accurate, that is my eyes and brain have compensated for reality and made me believe I have created the conditions necessary for this exercise but the reality was quite different. This indicates to me that the knowledge I think I possess is flawed - if I had looked at the histogram on the preview screen of my camera as I took these images I would have realised that I was not creating the correct conditions for the exercise. My belief that I was producing the correct conditions and my willingness to push ahead with the exercise have led to the mistake - the main learning I will take from this project is to consider more fully what is being asked in the brief and be less overly confident about where my level of knowledge sits. I also now realise I have much more to learn about the way my camera sees and records the world.

1 Perfect Exposure p.50, Michael Freeman
2 Photographers Eye p.110, Michael Freeman