One final act was to get the original seed pattern to 250,000, it having been the first rendered for canvas printing in late 2008, but long stuck just short of 250,000 by the limits of available computers. In the finish I took advantage of that pattern's symmetry and chopped off most of the lower half before running the top half and a sufficient buffer to 250,000, then running intermediate pixel data for those renderings through a simple filter to retain the top half and add a copy rotated 180°.

The other was to actually render the last third of the 500+ viable seeds run to 100,000, having done likewise for those run to 150,000 a couple of weeks earlier.