It is often desirable to reproduce a specific room-acoustic scene, e.g. a concert hall in a playback room, in such a way that the listener has a plausible and authentic spatial impression of the original sound source including the room acoustical properties. In this study a perceptually motivated approach for spatial audio reproduction is developed. This approach optimizes the spatial and monaural cues of the direct and reverberant sound separately. More specifically, the (monaural) spectral cues responsible for the timbre and the (binaural) interaural cross correlation (IACC) cues, responsible for the listener envelopment, were optimized in the playback room to restore the auditory impression of the recording room. The direct sound recorded close to the source is processed with an auditory motivated gammatone filterbank such that the spectral cues, ITD’s and ILD’s are comparable to the direct sound in the recording room. Additionally, the reverberant sound, which was recorded at two distant locations from the source, is played back via dipole loudspeakers. Due to the arrangement of the two dipole loudspeakers, only the diffuse sound field in the playback room is excited, therefore the spectral cues and the IACC of the reverberant sound field can be controlled independently to match the cues that were present in the recording room. As indicated by a preliminary listening test the applied optimization is perceptually similar to the reference signal and is generally preferred when compared to a conventional room-in-room reproduction.
展开▼
