I wonder if semiconductor Bloch equation solving can evaluate an exciton binding energy (G)? I think that inputting parameters are energy gap, dipole moment of band transition, and optical dielectric constant. I think that the formula G = \mu e^4/(\hbar (\epsilon_e \epsilon_0)^2) is not precise for binding energy evaluation, but Bethe-Salpeter equation solving is too expensive. That's why I am asking this question.