Rather than attempting to map an entire creek, I am considering representing a single section of a creek in detail – at a scale of 1 to 1. Furthermore, rather than attempting to represent the selected site via a single photograph, my interest is in employing a grid-based sampling approach. So a section of creek (probably the edge of a creek) of say 2.5 by 2.5 metres will be photographed via a set of overlapping square samples. These can then be stitched together to form an overall detailed map of the ground, which can ultimately be printed at the same size as the ground itself. The multiple samples enable a highly detailed overall composite to be produced. They would also seem to have the capacity to estrange the space – rather than representing a single visual perspectives the visual map emerges as a perverse and impossible composite. It is an analytical non-view.
In order to put together such photograph there is a need to enact a meticulous grid-based sampling process. I can imagine a complex tripod grid, with the camera moved from one axis point to another to capture each spatial sample. I may attempt something like this, but need to first get a clearer sense of what’s involved – so this afternoon I ran a less than rigorous process in my backyard with some climbing bolts arranged in a rough grid. The photograph below shows two versions of the grid. The first is just a single photograph from a single viewpoint. The second is a composite of 28 images; each image taken directly above a relevant bolt in my hazily conceived gridof 7 columns and 4 rows).
Single image at left, sampled composite at right
Apart from the apparently broken bolt at the upper right of the composite, there are a uncomfortable visual artefacts – the wall, for instance, is weirdly composited, seeming to lean inwards, partially obscuring the view of one bolt. Worth pursuing, but still unsure how to sample effectively?