direkt zum Inhalt springen

direkt zum Hauptnavigationsmenü

Sie sind hier

TU Berlin

Page Content

The Concert Hall and its Mediatization. Comparative Empirical Study of Binaurally Synthesized Natural and Electro-acoustically Transmitted Musical Performances

Today music is consumed mainly via media such as radio, CD, television, or online streaming, whereas the amount of live music consumption is at a near negligible fraction. However, fundamental perceptual differences in the consumption of live and recorded music have rarely been studied on an empirical basis, most probably due to challenges in empirical operationalization. However, recently developed advanced audiovisual rendering technologies provide realistic stimuli which furthermore allow for a direct comparison of live and mediatized situations while keeping the musical performance constant.

More specifically, data-based dynamic binaural synthesis was shown to provide plausible acoustic simulations, i.e. simulations which may not be discerned from acoustic reality when comparing to inner references of comparable real acoustic events. Moreover, this convincing acoustic simulation may be combined with photorealistic stereoscopic visual information, allowing for the first time conducting laboratory experiments that allow the immediate comparison of arbitrary media settings and live concerts while using most immersive stimuli.

To this end, a binaural measurement robot (FABIAN) which has been developed by the Audio Communication Group provides a fast and automated measurement of complete sets of so-called binaural room impulse responses (BRIRs). BRIRs have been acquired for different head orientations and different sound source positions in a concert hall, a 5-channel recording studio and a spatial audio laboratory equipped with a 56-channel loudspeaker array. Additionally, room impulse responses have been measured in the concert hall for standard microphone recording techniques. Finally, as acoustic stimuli several pieces performed by a professional string quartet have been recorded in the anechoic chamber of TU Berlin. Now, when convolving measured BRIRs or microphone impulse with the anechoic audio files a very convincing acoustic simulation can be achieved, all the more as the acoustic rendering may account in a natural fashion for head movements of the listener (dynamic binaural rendering). Furthermore, stereoscopic and conventional videos of performers have been recorded in the concert hall and may were used to produce visual stimuli for accompanying the live situation or augmenting the studio and spatial audio laboratory situation by respective music videos.

In a first empirical step these stimuli will be used to reveal the relevant qualitative features which differentiate between live and recorded music. Hence, in a listening test based on the Repertory Grid Technique a number of subjects will compare these stimuli while qualitative verbal descriptors will be elicited. Then, from these highly valid perceptual criteria a semantic differential will be constructed which then be used to test essential hypotheses on media consumption in a quantitative large-scale follow-up listening test.

This research project is funded by the German Research Foundation (DFG WE 4057/1-1)

Zusatzinformationen / Extras

Quick Access:

Schnellnavigation zur Seite über Nummerneingabe

Auxiliary Functions

This site uses Matomo for anonymized webanalysis. Visit Data Privacy for more information and opt-out options.