Same as for Fig. 14, but now for the observations of SN 2012au at 2269 d and using the he8p00 or he12p00 models with a magnetar power of 10⁴⁰ erg s⁻¹ and different levels of clumping. In each case, the model is scaled to match the same optical luminosity as inferred from observations. We also shift both models up by 2 × 10⁻¹⁸erg s⁻¹ cm⁻² Å to match the observed flux present between the emission lines in SN 2012au (the model predicts zero flux in those regions). [See Section 10.3 for discussion.]