Implemented some parallelisation in ASA calculation. It doesn't scale very well (could measure x2 speed-up with 4 CPUs in a Core2 Quad) but at least we have it, now this is starting to be better than NACCESS!