This work tackles the difficulties that appear with measuring- and analysis techniques for industrial centrifugal fans. In the made study we've tried to achieve data of a centrifugal fan in order to find a characterising fingerprint of the fan itself. Simply looking to the spectral information seems not to offer a detailed view of the underlying phenomenons. This due to several reasons: like the enormous impact of the generated wind and the limits of the techniques themselves. We've analysed the fan using soundpressure techniques and intensity techniques in anechoic and reverberant rooms respectively.We believe that a fingerprint requires at least both time and frequency information even if fan noise is considered stationary. The made measurements are analysed using wavelet analysis which offers a broader platform to characterise different relations between the mechanical construct and the noise. This by looking to self similarities, discontinuities, pure tones and short term evolution of the noise.