An apparent inadequacy of the velocity addition theorem in handling a Fizeau (like) experiment beyond the resonance frequence is shown not to be a true failing. The proof consists in showing that the invariance of the signal velocity is unrelated to its maximality. Next, Lorentz transformations which connect frames moving with supersignal velocity to those of subsignal velocity are derived, taking a cue from the acoustic case. Methods are also suggested for laboratory verification of the theory.