In this work two ways of processing controlled-source magnetotelluric (MT) data were tried and compared against each other. The aim was to evaluate the differences between a multivariate-based processing method to a bivariate processing method. The software KMSProTF represents conventional processing using bivariate and robust statistics. MsDEMPCA utilizes a multi-variate Criss-Cross regression scheme to improve the condition of the data-matrix before robustly decomposing it into principal components. Data from the FENICS-19 survey in northern Scandinavia was processed to transfer functions (TF) using the respective method. The TFs were visually interpreted in KMSProTF. There were no significant differences found between the methods. In addition a calibration between instruments was carried out, which caused an exclusion of parts of the data-set.