A Remark about dimension reduction for supremal functionals : the case with convex domain