Fig. 4

To construct initial conditions from artificially generated pinwheel maps, we first took such a pinwheel map with a certain scaling (upper left), then we chose a selection of excited orientations by means of a Gaussian. The width of the Gaussian gives the selection depth (upper right). After that, we masked the result spatially with another Gaussian distribution that is radially symmetric (lower right). The width of this Gaussian gives the third parameter, the size of the pattern. Finally the result is scaled, giving rise to the fourth parameter, we called the excess and added to the activator variable in the homogeneous state (lower left). The inhibitor variable is put into the homogeneous state