5th day.

By Juhani — 

Ideas for generative book design…

I’ve thought about using modules of some kind for ornamentation. These modules (of vector forms) could be positioned based on a hierarchy: 1st the canvas size is set (format) and all the text fields are positioned (content), then the remaining space is filled entirely by predefined modules of ornament. The modules also have a hierarchy of their own based on size: biggest modules are placed 1st, wherever they fit, followed by smaller and smaller modules. Possibly the shape and texture of the modules also change as they become smaller. The last module would be only a fraction of the 1st in size (something barely visible).

Alternatively, the modules could be closed line forms (“thread” loops) that are placed on a path that serves as some kind of spine for an ornamental border. The border could be mirrored in the x or y axis (or both). This would be somewhat similar to the example of generative book-covers linked in the previous post. The Faber covers are a bit primitive though, since there is no interaction between the modules. I would like to see the modules synthesized somehow: to create a fabric of ornament where no modules are actually placed over or underneath eachother, but instead tied together. Every “thread” has a modulating way of handling the intersections (over, under, over, under, &c.) becoming a part of the fabric. This results in a curious ambivalency of depth which is the special thing about the “knot” ornaments (illustrated in the previous post as well).

The “generative” concept could also be reversed. A 15th century book, for example, is printed with woodcuts (images) and movable metal type (text). Both materials are sensitive to degeneration; wearing out in constant use. (The former much more than the latter, I think.) A modern analogy for this could be photographs and jpeg compression. A jpeg file can be saved with a low quality, the low quality file opened, saved again with a lower quality, opened again, &c. The size of the file (amount of data) will reduce until some point and stop, but even after reaching the smallest size, the image keeps changing (the visual information degenerates). This kind of jpeg compression loop could be utilized for books that are printed one by one (print-on-demand): every book would have slightly different image content that degenerates in small steps until the end of the “print-run” (which is basically inifite).