Detection of aircraft targets varies as a function of target and masker temporal complexity, similarity, and realism

Abstract

The complexity of real-world auditory scenes makes it difficult to disentangle the factors that drive detectability of sounds within those scenes. However, many real-world applications of auditory perception research require the disentanglement of these factors – for example, to predict the effects of aircraft noise on the public depending on the ambient sound environment and the aircraft itself. The present study investigates effects of masker temporal complexity and similarity to targets on the detectability of helicopter target sounds. We investigate three levels of masker temporal complexity – urban sound scenes (high); synthetic versions of each scene that preserve overall statistics (medium); and noise of the same spectrum as the real scene (low). Target similarity was probed in a way that would occur in the world—as a function of the presence of other engine sounds. Each real masking sound could be comprised of only engine noise, engine sounds embedded in an urban scene, or a scene without engines. This provides maskers with three levels of target similarity. Combination with the three levels of temporal complexity allows us to probe interactions between temporal complexity and engine-related signal energy. Effects of target signal characteristics and interactions with signal to noise ratio will be discussed.