precisionChange lattice function now uses copyLane to remove need for temporary scalar objects, reducing register footprint and significantly improving performance