UV normalisation and T_eff fitting procedure for H35. Left panels: (a) the observed flux in units of 10⁻¹⁴ ergs cm⁻² s⁻¹ Å⁻¹, (b) the observed flux divided by the normalised cmfgen model (dark blue points, masked points are shown semi-transparent) and a fit through the non-masked points (red solid line), and (c) the normalised spectrum (dark blue solid line), the normalised, Ly-α corrected spectrum (blue dashed line) and the cmfgen model used for obtaining the normalisation model (solid orange line). The yellow shaded areas indicate the regions used for determining the goodness of fit of the iron pseudo-continuum of the model; roman numerals indicate the dominant iron ion in each region. Right panel: for each model T_eff and v_rad the dimensionless goodness of fit to the normalised iron pseudo-continuum of the data (defined as the inverse of the squared sum of the difference between the normalised fluxes of the model and the data in the iron continuum). At the peak lie the T_eff and v_rad that give the best fit: these are the values that are used for the final normalisation shown on the left.