Reproducibility of soundscape dimensions

The aim of this work was to investigate the validity of laboratory soundscape reproduction and the robustness of soundscape dimensional analysis. The soundscapes in four urban locations in Manchester (UK) were recorded with a soundfield microphone. The soundscapes were reproduced through an ambisonic 3D loudspeaker system in a semianechoic chamber. Fifteen listeners rated the reproduced soundscapes on nineteen elevenpoint semantic differential scales. Factor analysis revealed that four dimensions explained 65% of the variance in the judgements. The dimensions were: relaxation/calmness (41%), dynamics/vibrancy (10%), communication (7%) and spatiality (7%). This experiment can be compared with that of Kang (2007), who used very similar rating scales on outdoor soundwalks in two locations in Sheffield (UK). Kang found dimensions of relaxation (26%), communication (12%), spatiality (8%) and dynamics (7%). The dimensions from the two experiments load onto the semantic scales in a similar way. These results indicate that ambisonic reproduction of soundscapes gives similar results to field experiments, though with more variance explained. They also show that dimensional analysis of soundscape response is robust enough to produce similar results for different locations in different cities. It is suggested that the dimensions that result from this kind of experiment depend somewhat on the rating scales and questions used. Nevertheless, soundscapes and their perceptual dimensions both seem to be reproducible to a meaningful extent.