It is shown that in active region modelling of solar irradiance variations, care should be taken not to simply use average contrasts of faculae. This is particulrly the case if the model then ignores active region facular areas when the facular contrasts are low, as this then doubly devalues the contribution of faculae relative to sunspots.