A weak scattering model for tone haystacking