INTRODUCTION In the automotive industry, Noise, Vibration and Harshness (NVH) issues are important factors for the perceived quality of a product. Interior tire noise is an essential part of NVH. Tire noise is generated from the tire/road interaction and transferred into the car cabin through structure-borne and airborne paths. At low frequencies, typically less than 500 Hz [ 1], the structure-borne contribution dominates. Vibrations caused by the tire/road interaction are transferred through the hub to the suspension and chassis and radiate into the car cabin. At higher frequencies, the contribution from airborne tire noise dominates over the structure-borne tire noise. Although interior tire noise is an essential part of NVH and is highly related to comfort, it is often handled late in the car development process when there is little freedom for modifications of the car . For both car and tire manufacturers, it is desirable to predict tire noise early in the development process to allow time and cost efficient product development. This is achieved by moving field testing to indoor lab environments and using computer-aided engineering tools for early predictions of product performance and quality. A useful method to address NVH problems and to reduce field-testing is to combine recordings with measurements and/or simulations into auralizations. An example of a method to create auralizations is to filter the recordings through airborne and/or structure-borne transfer functions [2, 3]. Transfer functions are obtained from experimental measurements by a method such as Transfer Path Analysis (TPA), from numerical simulation methods such as Finite Element Modeling (FEM), or from methods combining these two techniques such as Component Mode Synthesis (CMS) and Frequency response function Based Substructuring (FBS) [4]. To create authentic structure-borne auralizations, the structural contributions from all six degrees of freedom (DOFs) - three translational and three rotational DOFs - have to be included. Due to factors such as measurement difficulty , complexity, time, and cost, rotational DOFs are often omitted and ignored [5, 6]. This results in simplified auralizations with varying magnitudes of audible errors. However, the required level of detail in the auralization depends on the stage in the development process. In an early development stage, audible errors and artifacts might be acceptable as long as the main character of the sound is realistic. In cases where auralizations are used for detailed psychoacoustic analysis, it is important to keep the auralizations perceptually equivalent to real sounds [ 7, 8]. What is considered important for preservation of a sound's main character may be application specific, but a basic requirement should be that listeners' preference ratings for a certain sound should not be altered due to errors and artifacts in the auralization [ 9]. The aim of this study was to investigate how different DOFs of hub forces and moments affect perceived tire noise in the cabin of a car. An auralization containing all six DOFs was compared with auralizations containing all but one DOF in a listening test to determine which DOFs are perceived as the most prominent in structure-borne interior tire noise. In addition, auralizations lacking Prominence of Different Directions of Hub Forces and Moments in Structure-Borne Tire Noise Magnus Löfdahl, Arne Nykänen, and Roger Johnsson Luleå University of Technology ABSTRACT In the automotive industry, tire noise is an important factor for the perceived quality of a product. A useful method to address such NVH problems is to combine recordings with measurements and/or simulations into auralizations. An example of a method to create structure-borne tire noise auralizations is to filter recordings of hub forces and moments through binaural transfer functions experimentally measured from the hub of the car to an artificial head in th