AbstractsCommunication

In order to develop a better understanding of the differences in the multichannel reproduction of solo piano music that result from the use of various popular multichannel microphone techniques, an experimental investigation employing both physical and perceptual measurements was undertaken. Four sophisticated multichannel microphone arrays were optimally placed for simultaneous capture of a series of four piano performances, so that resulting auditory imagery could be compared between otherwise identical performances. In an initial study, preference choices of two groups of listeners showed that not only the microphone techniques in use, but also differences in the musical content, contributed to the modulation of the listeners' preference. In order to predict the obtained preference of the listeners through the analysis of the binaural signals, differing versions of the musical programs were captured through the Head and Torso Simulator (HATS). The analysis showed that a combination of two electroacoustical measures, Ear Signal Incoherence (ESI) and Side Bass Ratio (SBR), was closely related to the obtained preference. The subsequent analysis revealed that the salient perceptual attributes were described by five bipolar pairs of adjectives, and the listener preference was mostly accounted by those attributes. However, the ESI and SBR were not associated with any of the salient attributes, which led another investigation to find the physical measures associated with the perceptual attributes. New physical measures were then derived from the analysis of binaurally captured multichannel reproduced piano music for WIDTH, BASS TIGHTNESS and SHARPNESS, respectively. The qu Afin de comprendre les caracteristiques latentes de l'enregistrement multicanaux de la musique reproduite d'un piano solo, quatre techniques sophistiquees de captation micro multicanaux ont ete deployees de facon optimale pour saisir les memes performances simultanement. Les resultats des echells de preferences chez les auditeurs en relation aux quatres techniques de captation multicanaux utilisees demontrent que non seulement les techniques de captation, mais aussi le contenu musical contribuent a faire moduler les preferences des auditeurs. Afin de predire les preferences obtenues des auditeurs par l'analyse des signaux binauraux, differentes versions des programmes musicaux furent captures e l'aide dun Simulateur Tete et Torse (HATS). L'analyse a demontre que la combinaison de deux mesures electroacoustiques, le Ear Signal Incoherence (ESI) et le Side Base Ratio (SBR), etait etroitement…