As previously discussed, I wrote code to generate the bisection, and to return an exactly bisected pictorial space, i.e. the area either side of the division is equal, but the algorithm is raw (un-optimized) and uses a trial-and-error method: it keeps generating a random-walked line until it generates one that is strictly normal.

Sequent 15 visualizes these two aspects. Firstly, it shows the definitive bisecting line that is strictly normal, viz an exact bisection. Secondly, it shows the other attempts that were not strict, viz they bisected the pictorial space but in an unequal manner.

Thoughts

I am not detailing the algorithm in use but rather the interim and final results of it.

There were many ways in which the data could have been presented. As I am thinking about overprinting, I decided to apply an aspect of that to this image.

The strictly normal line is shown in 100% M, and is painted normally.

The normal lines are overprinted in C, the opacity such that they total 100%.

The result is a great overprinting effect with frequently visited paths being more intense than others. The overprinted paths follow a fairly normal distribution so the random function works well.

Terminology

Strictly Normal Bisection — a line that bisects the pictorial space into exactly equal parts.

Normal Bisection — a line that bisects the pictorial space into unequal parts.

comments powered by Disqus