Synthetic Landscapes - created using Fortran
The images shown here were created using a Fortran program I have writen which builds up the image pixel by pixel, and allows manipulation of pixels in an ordered fashion using computational formulae with built in randomness. Each pixel of just about any photographic image on a computer is (basically) made up of a set of three values for each pixel (normally representing the amount of Red, Green, and Blue. The program manipulates these values.
I do this by creating colour washes, then adding details such as areas with random highlights or shapes, as well as shapes I have 'drawn' in terms of their co-ordinates which I can re-shape and fill with colour. The aim is for landscapes that look plausible, whilst being 'pure' in terms of their elements and composition.
While there are potentially no creative limits, there are practical limits to what a programming language of any type can create from scratch. I'm starting with what is simple and adding more complex elements when I think they will add to the potential for 'realism'. These images shown here are very small and lack detail, but an advantage of this approach is that very large images can be created.
I think the results give the appearance of printmaking more than photography. I think it's the sharp outlines of the features like birds that I put in that do this, and probably because in some cases the colours are more 'pure' than in photography.
I generally base the images on photographs that I've taken and use colours that I've taken from real photographs. The image on the left below is a photograph, and the one on the right is a synthesised version - actually the same 'pattern' as the larger image above, but without the oystercatcher.
I can easily change the colours and place objects in different places. The folowing image are the same 'pattern' but using colours sampled from different photographs and altering the locations, widths, heights and angles of the gannets.
Fortran is a computing language that was widely used in the 1970s, mainly for scientific uses. It is fairly basic (and quite similar to the computing language called "Basic") and is not designed for dynamic or interactive computing that most programs are capable of. But for creating and manipulating numbers it works well. I use it here because I learnt it when I was younger and don't feel that any other language I know would be better for my purposes. I also feel that as many aspects of computing become history it's good to keep old traditions alive!