In this talk, we present a re-analysis of different black hole merger gravitational wave events detected by the LIGO and Virgo interferometers with state-of-art phenomenological waveform models, IMRPhenomX and IMRPhenomT, which include higher spherical harmonics and spin precession. Due to their rapid and accurate evaluation of the waveforms, but also an automatisation of our Bayesian inference runs, we test the waveform model families, the improvements in the precession treatment, non-informative priors, and different sampler settings and codes. In most of the studied events, the influence of higher modes is small, unless it is a massive event. In this case, IMRPhenomT further improves the fit to the data over IMRPhenomX owing to dropping the SPA approximation and other improvements in the waveform modeling. The prior choices also play an important role in challenging short signals.