An improved method of calculating finite range corrections in two-nucleon transfer reactions