The filtration of a fiber suspension has been studied experimentally. Typical application where pressure filtration occurs are: papermaking, air cleaners, production of composite materials, etc. In particular, in papermaking, the quality of the final product depends on the fiber orientation and mass distribution in the filtered materials Micro-variations of there properties can strongly affect the quality of the final product and they can occur during filtration, thus it is important to predict how this can happen. However, this is not an easy task, first because the filtered cake is a non-homogeneous compressible porous media, second because the filtration flow is non-stationary, since the cake is continuously evolving in time. Therefor in this work we focus on the filtration flow through formed steady fiber networks. For each grammage (i.e. mass of fibers per unit area), we simultaneously measure the pressure drop across the network and velocity field on top and below the fiber network using Particle Image Velocimetry (PIV). Compression of the fiber network can also be extracted form the PIV images normalized filtration resistance was found to be decreasing with increasing network thickness, as well as network compressibility. From the PIV data the influence of the formed fiber network on the flow field was analyzed and characteristic scales of the flow structure are quantified.