Instead of using eta=-delta, use zeta=-(delta+1/2) to represent delta. This variant only needs at most 590 iterations for 256-bit inputs rather than 724 (by convex hull bounds analysis).
Instead of using eta=-delta, use zeta=-(delta+1/2) to represent delta. This variant only needs at most 590 iterations for 256-bit inputs rather than 724 (by convex hull bounds analysis).