This series is based on almost the same program as the "Complex Mapping" series, except that this time the source map is not a regular, artificially created pattern. It is an original photograph. (The algorithms were inspired by the book Creating Symmetry discussed on the Books page, but I have not exactly followed the maths suggested there: I find that images are often more interesting when there are departures from perfect symmetry.) In some cases the transformation is so extreme that we are doing little more than using the original photograph to define a palette of colours which are then used to produce an abstract pattern.
The processed images are shown first, followed by the original on which they were based.
In my opinion some subjects work better than others. My daughter Katie immediately commented that using a subject, such as a flower where you expect some original symmetry does not produce such interesting transformations. We agreed that the transformation of the trees in snow was successful. This may be partly due to the fractal complexity of the original subject, and partly due to the (rather loose) reference to the symmetry of snow crystals in some of the transformations. The image of the tree at sunset also created some results that I rather liked. I also liked the river delta, where the transformed image emphasises the essential morphology of the delta. I also felt that the modified images of The Bridge of Sighs at St Johns in Cambridge has some echos of Escher's impossible buildings. Most surprising was the transformations of the corn-flower, producing results that appeared to turn it into a spider's web. Strangely disturbing.
I also found that some photographs that have less striking subject - for example, the texture of vegetation or small waves on the sea - can produce interesting effects when one imposes some degree of symmetry.
These are really just initial experiments - an investigation of which type of subjects are likely to produce the most interesting results. See my page "Images from Complex Maps" for more discussion.
