Research project PlaNaS

Natural sound sources such as the human voice, musical instruments, and everyday sounds shape our auditory environment. In virtual acoustic environments, for example in applications of augmented and extended reality, such sound sources should be represented as realistically as possible. However, the spatial radiation characteristics of natural sound sources are significantly more complex than those of loudspeakers commonly used in many research studies.

In contrast to loudspeakers, which can be characterized by time-invariant directivity patterns and a well-defined acoustic center, natural sound sources exhibit spatially extended and often time-varying radiation behavior. In addition, movements of the sound source within the environment, as well as changes in orientation, influence the resulting sound radiation.

Many current studies comparing real and virtual acoustic reproduction rely on highly simplified sound source models. Since the differences compared to natural sound sources are clearly perceptible, they significantly affect the results of psychoacoustic investigations. This particularly concerns the assessment of authenticity (the complete sensory agreement between reality and simulation) and plausibility (the consistency of the presentation with the listener’s inner reference).

This is where the PlaNaS research project comes in. Its objective is to develop methods for reproducing the direction-dependent and spatially distributed sound radiation of natural sound sources in virtual acoustic environments. Building on this, psychoacoustic experiments are conducted to investigate which physical and perceptual factors influence plausibility and what level of reproduction accuracy is required.

A particular focus of the project lies on:

• modeling the complex, time-varying directivity patterns of natural sound sources,
• consideration of source extent and movement dynamics,
• as well as the further development and adaptation of established test paradigms for plausibility evaluation.

By combining advanced signal processing methods for the realization of virtual acoustic environments with psychoacoustic experiments, the project contributes to bringing these environments closer to real listening situations and improving their application in future technologies.