TECHNICAL CHALLENGES IN PREPARATION OF LISTENING TESTS Lukáš Zelem 1 , Vojtech Chmelík 2 , Herbert Müllner 3 , Rudolph Exel 4 and Monika Rychtáriková 5 1 Department of Building Structures, Slovak University of Technology in Bratislava, Faculty of Civil Engineering, Radlinského 11, 813 68 Bratislava, Slovakia, Email: zelem.lukas@stuba.sk 2 Department of Building Structures, Slovak University of Technology in Bratislava, Faculty of Civil Engineering, Radlinského 11, 813 68 Bratislava, Slovakia, Email: vojtech.chmelik@stuba.sk 3 Fachbereich Akustik & Bauphysik,TGM – Die Schule der Technik, Wexstraße 19 – 23, A-1200 Wien, Austria, Email: herbert.muellner@tgm.ac.at 4 Unternehmensberatung Rudolf Exel, A-7423 Grafenschachen 343, Graz, Austria, Email: beratung @ exel.at 5 KU Leuven, Faculty of Architecture, Hoogstraat 51, 9000 Gent/ Paleizenstraat 65, 1030 Brussel, Belgium, Email: monika.rychtarikova@kuleuven.be ABSTRACT This article is focused on a process of preparation of psychoacoustic tests for testing a subjective perception of airborne sound insulation from a point of view of technical support, such as software, hardware and listening procedures. Typical challenges occur in relation to optimization of laboratory environment for listening test experiments, choice of a convenient sound reproduction systems, users friendly interface, and also decision on a number of necessary participants, length of the tests and proper analysis method. All the mentioned issues have an impact on a global accuracy of experiments and will be discussed part by part in the paper. INTRODUCTION The light weight walls in comparison with traditional masonry or concrete walls have very different sound insulation spectra. Heavy-weight walls typically perform better at low frequencies and insulate well bass tones, whereas light-weight double walls, that use a mass-spring-mass system perform better at middle frequency ranges and insulate very well e.g. human speech. For this reason, it is very difficult to choose one general single number quantity, that could act as an objective parameter in assessment and rating of sound insulation. Previous studies have already shown, that to resolve this problem, listening tests are necessary [1], [2]. Other researcher has pointed out, that technical aspects and chosen psychoacoustic methods might influence the research results [3]. Therefore, this paper aims at discussion of technical challenges in preparation of listening tests that should serve to find out, which descriptor (RW, RW (living), RW,modif. noise, Roptimal, RW (20 phon), RW (30 phon), RW (40 phon), RW (TGM), RA, RA,tr, RW + C, Rspeech, Rtrafiic, R + Ctr), corresponds to subjective assessment of sound insulation the best. STIMULI Sound stimuli were prepared based on typical sounds in households [4] [5]. These stimuli were filtered according to sound insulation spectra R of chosen walls. In order to optimize a test, three types of test: (1)