In order to decide the preferred focal lenght and point of view for this assignment, I took a series of pictures with various cameras and lens combinations over several days. All the lenses used were primes, and ranged from 28mm to 100mm equivalent in full frame format.
I started with a 18mm prime lens on an APS camera (equivalent to about 28mm). This allowed me to include a reasonable amount of background. I was using a moderate aperture of f/5.6 (equivalent to a depth of f/8 in full frame), which considering the fact that most of the shoots I took with this were at relativelly long distances, allowed me to have most of the people in the scene in focus. I quickly decided this focal length was not going to work for what I wanted, because the distance with the subjects was too great, and I did not want to be too close to my subjects to avoid perturbing their behaviour. This focal lenght was also only going to work with large crowds, as otherwise there were going to be too many gaps, as in the pictute below
I subsequently moved to more normal angles and tried both a 40mm equivalent and 43mm prime lenses. These were tested again at the moderate aperture of f/8 (equivalent). As the focal lenghts increased from wide-angle, maintaining moderate apertures while focusing in nearer subjects resulted in a “layering” effect, with objects further away from the focus point becoming slightly blurred. I tried both portrait and landscape orientations with these lenses
The normal focal lengths were working better for me in terms of the prominence of the crowds than the wideangle, but I was still needed to be too close to the subjects in order to achieve the desired proportionality between subject and background. Pictures 6 and 7, for instance, were taken almost among the crowd. This was not what I wanted, as I intended to portray the subjects as an outsider.
At this point I decided to take all subsequent pictures on a landscape orientation. Crowds seem to spread naturally on a horizontal way and many of the pictures I took on a vertical orientation looked strangely unatural to me, as if they had too much unnecessary information at the top and on the bottom (compare pictures 2 and 3 above, for instance). Besides, I am more naturally inclined to take vertical shots, so sticking with the landscape orientation was also in part to challenge myself to look at a scene differently.
I finally moved to telephoto prime lenses. I used both 90mm (equiv) and 100mm lenses, but reduced the aperture to the equivalent of f/5.6. The compression of the view combined with moderate to wide aperture acentuated the “layering” effect, particularly when the subjects were relativelly close.
At this stage I also tried different points of view, both elevated (eg pictures 1 and 11) and from below (eg picture 10). Most of the pictures, however, were taken straight at eye level, and in the end I felt that this resulted in the less “menacing” representation of crowds, so I decided to stick to this point of view.
In the end, I believe the best approach was for me to carry out the shooting sessions using a telephoto lens at a moderatelly wide aperture. This enabled me to maintan crowds as the focal point of the image while keeping some elements of background for reference. More importantly, it enabled me to maintain a reasonable distance from the crowd and remain as an anonimous outsider most of the time, in satisfaction of my previoulsly stated objective of trying to distill the social identity elements of the crowd.